diff --git a/source/bibs/main.bib b/source/bibs/main.bib
index 3e22da1..203bf4e 100755
--- a/source/bibs/main.bib
+++ b/source/bibs/main.bib
@@ -789,14 +789,15 @@ @article{rogner_assessment_1997
 }
 
 
-@article{riahi_shared_2016,
+@article{riahi_shared_2017,
 	title = {The {Shared} {Socioeconomic} {Pathways} and their {Energy}, {Land} {Use}, and {Greenhouse} {Gas} {Emissions} {Implications}},
-	volume = {in press},
+	volume = {42},
+	pages = {153-168},
 	doi = {10.1016/j.gloenvcha.2016.05.009},
 	journal = {Global Environmental Change},
 	author = {Riahi, Keywan and Vuuren, Detlef P. van and Kriegler, Elmar and Edmonds, Jae and O’Neill, Brian and Fujimori, Shinichiro and Bauer, Nico and Calvin, Katherine and Dellink, Rob and Fricko, Oliver and Lutz, Wolfgang and Popp, Alexander and Cuaresma, Jesus Crespo and KC, Samir and Leimbach, Marian and Jiang, Leiwen and Kram, Tom and Rao, Shilpa and Emmerling, Johannes and Ebi, Kristie and Hasegawa, Tomoko and Havlik, Petr and Humpenoder, Florian and Silva, Lara Aleluia Da and Smith, Steve and Stehfest, Elke and Bosetti, Valentina and Eom, Jiyong and Gernaat, David and Masui, Toshihiko and Rogelj, Joeri and Strefler, Jessica and Drouet, Laurent and Krey, Volker and Luderer, Gunnar and Harmsen, Mathijs and Takahashi, Kiyoshi and Baumstark, Lavinia and Doelman, Jonathan and Kainuma, Mikiko and Klimont, Zbigniew and Marangoni, Giacomo and Lotze-Campen, Hermann and Obersteiner, Michael and Tabeau, Andrzej and Tavoni, Massimo},
 	url = {http://pure.iiasa.ac.at/13280/},
-	year = {2016}
+	year = {2017}
 }
 
 
@@ -849,21 +850,23 @@ @article{pietzcker_solar_2014
 }
 
 
-@article{eurek_wind_2016,
+@article{eurek_wind_2017,
 	title = {An improved global wind resource estimate for integrated assessment models},
 	author = {Eurek, K. and Sullivan, P. and Gleason, M. and Hettinger, D. and Heimiller, D.M. and Lopez, A.},
 	journal = {Energy Economics},
-	volume = {In Review},
-	year = {2016}
+	volume = {64},
+	pages = {552-567},
+	year = {2017}
 }
 
 
-@article{fricko_marker_2016,
+@article{fricko_marker_2017,
 	title = {The marker quantification of the shared socioeconomic pathway 2: a middle-of-the-road scenario for the 21st century},
-	volume = {In press},
+	volume = {42},
+	pages = {251-267},
 	journal = {Global Environmental Change},
 	author = {Fricko, Oliver and Havlik, Petr and Rogelj, Joeri and Klimont, Zbigniew and Gusti, Mykola and Johnson, Nils and Kolp, Peter and Strubegger, Manfred and Valin, Hugo and Amann, Markus and Ermolieva, Tatiana and Forsell, Nicklas and Herrero, Mario and Heyes, Chris and Kindermann, Georg and Krey, Volker and McCollum, David L. and Obersteiner, Michael and Pachauri, Shonali and Rao, Shilpa and Schmid, Erwin and Schoepp, Wolfgang and Riahi, Keywan},
-	year = {2016}
+	year = {2017}
 }
 
 
@@ -1544,7 +1547,6 @@ @article{loew_2016
 	publisher = {IOP Publishing}
 }
 
-
 @article{Raptis_2016_powerplant_data,
 	author = {Raptis, Catherine E. and Pfister, Stephan},
 	title = {{Global freshwater thermal emissions from steam-electric power plants with once-through cooling systems}},
@@ -1555,12 +1557,13 @@ @article{Raptis_2016_powerplant_data
 	year = {2016}
 }
 
-@article{huppmann_2019_MESSAGEix,
-   author = {Huppmann, Daniel and Gidden, Matthew and Fricko, Oliver and Kolp, Peter and Orthofer, Clara and Pimmer, Michael and Kushin, Nikolay and Vinca, Adriano and Mastrucci, Alessio and Riahi, Keywan and Krey, Volker},
-   title = {{The MESSAGEix Integrated Assessment Model and the ix modeling platform (ixmp): An open framework for integrated and cross-cutting analysis of energy, climate, the environment, and sustainable development}},
-   journal = {Environmental Modelling \& Software},
-   volume = {112},
-   pages = {143-156},
-   doi = {https://doi.org/10.1016/j.envsoft.2018.11.012},
-   year = {2019}
-}
+@article{huppmann_message_2019,
+	author = {Huppmann, Daniel and Gidden, Matthew and Fricko, Oliver and Kolp, Peter and Orthofer, Clara and Pimmer, Michael and Kushin, Nikolay and Vinca, Adriano and Mastrucci, Alessio and Riahi, Keywan and Krey, Volker},
+	title = {The MESSAGEix Integrated Assessment Model and the ix modeling platform (ixmp): An open framework for integrated and cross-cutting analysis of energy, climate, the environment, and sustainable development},
+	journal = {Environmental Modelling & Software},
+	volume = {112},
+	pages = {143–156},
+	year = {2019},
+	url = {https://www.sciencedirect.com/science/article/pii/S1364815218302330 },
+	type = {Journal Article}
+}
\ No newline at end of file
diff --git a/source/climate/index.rst b/source/climate/index.rst
index bf108ad..3a53cfd 100755
--- a/source/climate/index.rst
+++ b/source/climate/index.rst
@@ -9,6 +9,6 @@ like air pollutants, together with consistent projections of radiative forcing,
 MAGICC is most commonly used in a deterministic setup (Meinshausen et al., 2011b :cite:`meinshausen_rcp_2011`), but also a probabilistic setup (Meinshausen et al., 2009 
 :cite:`meinshausen_greenhouse-gas_2009`) is available which allows to estimate the probabilities of limiting warming to below specific temperature levels given a specified emissions 
 path (Rogelj et al., 2013a :cite:`rogelj_2020_2013`; Rogelj et al., 2013b :cite:`rogelj_probabilistic_2013`; Rogelj et al., 2015 :cite:`rogelj_mitigation_2015`). Climate feedbacks on 
-the global carbon cycle are accounted for through the interactive coupling of the climate model and a range of gas-cycle models. (Fricko et al., 2016 :cite:`fricko_marker_2016`)
+the global carbon cycle are accounted for through the interactive coupling of the climate model and a range of gas-cycle models. (Fricko et al., 2017 :cite:`fricko_marker_2017`)
 
 For more information about the model, see `www.magicc.org <http://www.magicc.org/>`_.
diff --git a/source/emissions/globiom/index.rst b/source/emissions/globiom/index.rst
index 7e9d7d5..71839d4 100644
--- a/source/emissions/globiom/index.rst
+++ b/source/emissions/globiom/index.rst
@@ -5,21 +5,42 @@ Emissions from land (GLOBIOM)
 
 Crop sector emissions
 ~~~~
-Crop emissions sources accounted in GLOBIOM are N2O fertilization emissions, from synthetic fertilizer and from organic fertilizers, as well as CH4 methane emissions from rice cultivation. Synthetic fertilizers are calculated on a Tier 1 approach, using the information provided by EPIC on the fertilizer use for each management system at the Simulation Unit level and applying the emission factor from IPCC AFOLU guidelines. Synthetic fertilizer use is therefore built in a bottom up approach, but upscaled to the International Fertilizer Association statics on total fertilizer use per crop at the national level for the case where calculated fertilizers are found too low at the aggregated level. This correction ensures a full consistency with observed fertilizer purchases. In the case of rice, only a Tier 1 approach was applied, with a simple formula where emissions are proportional to the area of rice cultivated. Emission factor is taken from EPA (EPA 2012 :cite:`environmental_protection_agency_epa_US_2012`).
+Crop emissions sources accounted in GLOBIOM are N2O fertilization emissions, from synthetic fertilizer and from organic fertilizers, as well as CH4 methane emissions from rice cultivation. 
+Synthetic fertilizers are calculated on a Tier 1 approach, using the information provided by EPIC on the fertilizer use for each management system at the Simulation Unit level and applying 
+the emission factor from IPCC AFOLU guidelines. Synthetic fertilizer use is therefore built in a bottom up approach, but upscaled to the International Fertilizer Association statics on total 
+fertilizer use per crop at the national level for the case where calculated fertilizers are found too low at the aggregated level. This correction ensures a full consistency with observed fertilizer purchases. 
+In the case of rice, only a Tier 1 approach was applied, with a simple formula where emissions are proportional to the area of rice cultivated. Emission factor is taken from EPA 
+(2012) :cite:`environmental_protection_agency_epa_US_2012`.
 
 Livestock emissions
 ~~~~
-In GLOBIOM, the following emission accounts were assigned to livestock directly: CH4 from enteric fermentation, CH4 and N2O from manure management, and N2O from excreta on pasture (N2O from manure applied on cropland is reported in a separate account linked to crop production). In brief, CH4 from enteric fermentation is a simultaneous output of the feed-yield calculations done with the RUMINANT model, as well as nitrogen content of excreta and the amount of volatile solids. The assumptions about proportions of different manure management systems, manure uses, and emission coefficients are based on detailed literature review. A detailed description of how these coefficients have been determined including the literature review is provided in (Herrero, Havlik et al. 2013 :cite:`herrero_global_2013`).
+In GLOBIOM, the following emission accounts were assigned to livestock directly: CH4 from enteric fermentation, CH4 and N2O from manure management, and N2O from excreta on pasture 
+(N2O from manure applied on cropland is reported in a separate account linked to crop production). In brief, CH4 from enteric fermentation is a simultaneous output of the feed-yield 
+calculations done with the RUMINANT model, as well as nitrogen content of excreta and the amount of volatile solids. The assumptions about proportions of different manure management systems, 
+manure uses, and emission coefficients are based on detailed literature review. A detailed description of how these coefficients have been determined including the literature review is provided 
+in (Herrero et al., 2013 :cite:`herrero_global_2013`).
 
 Land use change emissions
 ~~~~
-Land use change emissions are computed based on the difference between initial and final land cover equilibrium carbon stock. For forest, above and below-ground living biomass carbon data are sourced from (Kindermann, Obersteiner et al. 2008    :cite:`kindermann_global_2008`), where geographically explicit allocation of the carbon stocks is provided. The carbon stocks are consistent with the 2010 Forest Assessment Report (FAO 2010 :cite:`food_and_agricultural_organization_fao_global_2010`). Therefore, the emission factors for deforestation are in line with those of FAO. Additionally, carbon stock from grasslands and other natural vegetation is also taken into account using the above and below ground carbon from the biomass map from (Ruesch and Gibbs 2008 :cite:`ruesch_new_ipcc_2008`). When forest or natural vegetation is converted into agricultural use, it is considered in this approach that all below and above ground biomass is released in the atmosphere. However, the following are not accounted for: litter, dead wood and soil organic carbon.
+Land use change emissions are computed based on the difference between initial and final land cover equilibrium carbon stock. For forest, above and below-ground living biomass carbon data are sourced from 
+Kindermann et al. (2008) :cite:`kindermann_global_2008`, where geographically explicit allocation of the carbon stocks is provided. The carbon stocks are consistent with the 2010 Forest Assessment Report 
+(FAO, 2010 :cite:`food_and_agricultural_organization_fao_global_2010`). Therefore, the emission factors for deforestation are in line with those of FAO. Additionally, carbon stock from grasslands and other 
+natural vegetation is also taken into account using the above and below ground carbon from the biomass map from (Ruesch and Gibbs, 2008 :cite:`ruesch_new_ipcc_2008`). 
+When forest or natural vegetation is converted into agricultural use, it is considered in this approach that all below and above ground biomass is released in the atmosphere. 
+However, the following are not accounted for: litter, dead wood and soil organic carbon.
 
 Comparison with other literature
 ~~~~
-In order to put the numbers in perspective with other sources they were compared with FAO (Tubiello, Salvatore et al. 2013 :cite:`tubiello_faostat_2013`) where a simple but transparent approach is used, largely relying on FAOSTAT activity numbers and IPCC Tier 1 emission coefficients (see :numref:`tab-globff`).
+In order to put the numbers in perspective with other sources they were compared with FAO (Tubiello et al., 2013 :cite:`tubiello_faostat_2013`) where a simple but transparent approach is used, largely relying on FAOSTAT 
+activity numbers and IPCC Tier 1 emission coefficients (see :numref:`tab-globff`).
 
-The 2000 data for crops are overall about 11% higher than Tubiello et al., mainly because of rice where the data are closer to EPA (EPA 2012 :cite:`environmental_protection_agency_epa_US_2012`) which is higher than Tubiello et al. For livestock, it is by some 18% lower than Tubiello et al. So in total there is about 10% GHG emissions less in 2000 than the values reported. The year 2010 is already the result of simulations and hence may be interesting to compare with the data. In order to facilitate the comparison, the columns e), f) and g) in Table 1 are3 included. Columns e) and f) compare GLOBIOM data for 2000 and projections for 2010 respectively, with numbers reported by Tubiello et al. Column g) compares the relative change in emissions between 2000 and 2010 from these two sources (1.00 would indicate the same relative change in GLOBIOM and in Tubiello et al.). It is apparent that the relative change in total agricultural emissions in GLOBIOM is the same as the development reported by Tubiello et al. – an increase by 11%. The behavior of GLOBIOM is over this period very close to the reported trends also at the level of individual accounts. The only exception is emissions from manure management where the relative change projected in GLOBIOM is by 13% higher than the relative change observed in Tubiello's numbers. 
+The 2000 data for crops are overall about 11% higher than Tubiello et al., mainly because of rice where the data are closer to EPA (EPA 2012 :cite:`environmental_protection_agency_epa_US_2012`) which is higher than 
+Tubiello et al. For livestock, it is by some 18% lower than Tubiello et al. So in total there is about 10% GHG emissions less in 2000 than the values reported. The year 2010 is already the result of simulations 
+and hence may be interesting to compare with the data. In order to facilitate the comparison, the columns e), f) and g) in Table 1 are3 included. Columns e) and f) compare GLOBIOM data for 2000 and projections for 
+2010 respectively, with numbers reported by Tubiello et al. Column g) compares the relative change in emissions between 2000 and 2010 from these two sources (1.00 would indicate the same relative change in GLOBIOM 
+and in Tubiello et al.). It is apparent that the relative change in total agricultural emissions in GLOBIOM is the same as the development reported by Tubiello et al. – an increase by 11%. The behavior of GLOBIOM 
+is over this period very close to the reported trends also at the level of individual accounts. The only exception is emissions from manure management where the relative change projected in GLOBIOM is by 13% higher 
+than the relative change observed in Tubiello's numbers. 
 
 .. _tab-globff:
 .. list-table:: Comparison of agricultural GHG emissions from GLOBIOM and from FAO for the years 2000 and 2010
diff --git a/source/emissions/message/ghgs.rst b/source/emissions/message/ghgs.rst
index fc05aaa..1ce78ce 100644
--- a/source/emissions/message/ghgs.rst
+++ b/source/emissions/message/ghgs.rst
@@ -2,8 +2,14 @@ GHGs
 ===========
 Carbon-dioxide (CO2)
 ---------------------
-The MESSAGE model includes a detailed representation of energy-related and land-use CO2 emissions (Riahi and Roehrl, 2000 :cite:`riahi_greenhouse_2000`; Riahi, Rubin et al., 2004 :cite:`riahi_prospects_2004`; Rao and Riahi, 2006 :cite:`rao_role_2006`; Riahi et al., 2011 :cite:`riahi_rcp_2011`). Energy related CO2 mitigation options include technology and fuel shifts; efficiency improvements; and carbon capture. A number of specific mitigation technologies are modeled bottom-up in MESSAGE with a dynamic representation of costs and efficiencies. MESSAGE also includes a detailed representation of carbon capture and sequestration from both fossil fuel and biomass combustion. Land-use CO2 was previously represented using methodology documented in Riahi et al. (2007) :cite:`riahi_scenarios_2007` but is currently updated based on information from the GLOBIOM model.
+The MESSAGE model includes a detailed representation of energy-related and land-use CO2 emissions (Riahi and Roehrl, 2000 :cite:`riahi_greenhouse_2000`; Riahi, Rubin et al., 2004 :cite:`riahi_prospects_2004`; 
+Rao and Riahi, 2006 :cite:`rao_role_2006`; Riahi et al., 2011 :cite:`riahi_rcp_2011`). Energy related CO2 mitigation options include technology and fuel shifts; efficiency improvements; and carbon capture. 
+A number of specific mitigation technologies are modeled bottom-up in MESSAGE with a dynamic representation of costs and efficiencies. MESSAGE also includes a detailed representation of carbon capture 
+and sequestration from both fossil fuel and biomass combustion. Land-use CO2 was previously represented using methodology documented in Riahi et al. (2007) :cite:`riahi_scenarios_2007` but is currently 
+updated based on information from the GLOBIOM model.
 
 Non-CO2 GHGs
 -------------------
-MESSAGE includes a representation of non-CO2 GHGs (CH4, N2O, HFCs, SF6, PFCs) mandated by the Kyoto Protocol (Rao and Riahi, 2006 :cite:`rao_role_2006`). Included is a representation of emissions and mitigation options from both energy related processes as well as non-energy sources like livestock, municipal solid waste disposal, manure management, fertilizer use, rice cultivation, wastewater, and crop residue burning.
+MESSAGE includes a representation of non-CO2 GHGs (CH4, N2O, HFCs, SF6, PFCs) mandated by the Kyoto Protocol (Rao and Riahi, 2006 :cite:`rao_role_2006`). 
+Included is a representation of emissions and mitigation options from both energy related processes as well as non-energy sources like livestock, municipal solid waste disposal, 
+manure management, fertilizer use, rice cultivation, wastewater, and crop residue burning.
diff --git a/source/emissions/message/index.rst b/source/emissions/message/index.rst
index 2710217..83ec805 100644
--- a/source/emissions/message/index.rst
+++ b/source/emissions/message/index.rst
@@ -61,4 +61,12 @@ MESSAGE includes a representation of non-CO2 GHGs (CH4, N2O, HFCs, SF6, PFCs) ma
 
 Air pollution
 ~~~~~~~~~~~~~
-Air pollution implications are derived with the help of the GAINS (Greenhouse gas-Air pollution INteractions and Synergies) model. GAINS allows for the development of cost-effective emission control strategies to meet environmental objectives on climate, human health and ecosystem impacts until 2030 (Amann et al., 2011 :cite:`amann_cost-effective_2011`). These impacts are considered in a multi-pollutant context, quantifying the contributions of sulfur dioxide (SO2), nitrogen oxides (NOx), ammonia (NH3), non-methane volatile organic compounds (VOC), and primary emissions of particulate matter (PM), including fine and coarse PM as well as carbonaceous particles (BC, OC). As a stand-alone model, it also tracks emissions of six greenhouse gases of the Kyoto basket with exception of NF3. The GAINS model has global coverage and holds essential information about key sources of emissions, environmental policies, and further mitigation opportunities for about 170 country-regions. The model relies on exogenous projections of energy use, industrial production, and agricultural activity for which it distinguishes all key emission sources and several hundred control measures. GAINS can develop finely resolved mid-term air pollutant emission trajectories with different levels of mitigation ambition (Cofala et al., 2007 :cite:`cofala_scenarios_2007`; Amann et al., 2013 :cite:`amann_regional_2013`). The results of such scenarios are used as input to global IAM frameworks to characterize air pollution trajectories associated with various long-term energy developments (see further for example Riahi et al., 2012 :cite:`riahi_chapter_2012`; Rao et al., 2013 :cite:`rao_better_2013`; Fricko et al., 2016 :cite:`fricko_marker_2016`).
+Air pollution implications are derived with the help of the GAINS (Greenhouse gas-Air pollution INteractions and Synergies) model. GAINS allows for the development of cost-effective emission control strategies to 
+meet environmental objectives on climate, human health and ecosystem impacts until 2030 (Amann et al., 2011 :cite:`amann_cost-effective_2011`). These impacts are considered in a multi-pollutant context, 
+quantifying the contributions of sulfur dioxide (SO2), nitrogen oxides (NOx), ammonia (NH3), non-methane volatile organic compounds (VOC), and primary emissions of particulate matter (PM), including fine 
+and coarse PM as well as carbonaceous particles (BC, OC). As a stand-alone model, it also tracks emissions of six greenhouse gases of the Kyoto basket with exception of NF3. The GAINS model has global 
+coverage and holds essential information about key sources of emissions, environmental policies, and further mitigation opportunities for about 170 country-regions. The model relies on exogenous projections 
+of energy use, industrial production, and agricultural activity for which it distinguishes all key emission sources and several hundred control measures. GAINS can develop finely resolved mid-term air pollutant 
+emission trajectories with different levels of mitigation ambition (Cofala et al., 2007 :cite:`cofala_scenarios_2007`; Amann et al., 2013 :cite:`amann_regional_2013`). The results of such scenarios are used as 
+input to global IAM frameworks to characterize air pollution trajectories associated with various long-term energy developments 
+(see further for example Riahi et al., 2012 :cite:`riahi_chapter_2012`; Rao et al., 2013 :cite:`rao_better_2013`; Fricko et al., 2017 :cite:`fricko_marker_2017`).
diff --git a/source/energy/conversion/electricity.rst b/source/energy/conversion/electricity.rst
index c662c9a..e3e52ce 100755
--- a/source/energy/conversion/electricity.rst
+++ b/source/energy/conversion/electricity.rst
@@ -74,14 +74,14 @@ Most thermal power plants offer the option of coupled heat production (CHP, see
 .. figure:: /_static/costind-thermo.png
    :width: 700px
 
-   Cost indicators for thermoelectric power-plant investment (Fricko et al., 2016 :cite:`fricko_marker_2016`). 
+   Cost indicators for thermoelectric power-plant investment (Fricko et al., 2017 :cite:`fricko_marker_2017`). 
 
-In :numref:`fig-ther`, the black ranges show historical cost ranges for 2005. Green, blue, and red ranges show cost ranges in 2100 for SSP1, SSP2, and SSP3, respectively (see description of the :ref:`narratives`). Global values are represented by solid ranges. Values in the global South are represented by dashed ranges. The diamonds show the costs in the “North America” region (Fricko et al., 2016 :cite:`fricko_marker_2016`).
+In :numref:`fig-ther`, the black ranges show historical cost ranges for 2005. Green, blue, and red ranges show cost ranges in 2100 for SSP1, SSP2, and SSP3, respectively (see description of the :ref:`narratives`). Global values are represented by solid ranges. Values in the global South are represented by dashed ranges. The diamonds show the costs in the “North America” region (Fricko et al., 2017 :cite:`fricko_marker_2017`).
 
 .. _fig-nonth:
 .. figure:: /_static/costind-nonthermo.png
    :width: 700px
 
-   Cost indicators for non-thermoelectric power-plant investment (Fricko et al., 2016 :cite:`fricko_marker_2016`). Abbreviations: CCS – Carbon Capture and Storage; IGCC – Integrated gasification combined cycles; ST – Steam turbine; CT – Combustion turbine; CCGT – Combined cycle gas turbine 
+   Cost indicators for non-thermoelectric power-plant investment (Fricko et al., 2017 :cite:`fricko_marker_2017`). Abbreviations: CCS – Carbon Capture and Storage; IGCC – Integrated gasification combined cycles; ST – Steam turbine; CT – Combustion turbine; CCGT – Combined cycle gas turbine 
    
-In :numref:`fig-nonth`, the black ranges show historical cost ranges for 2005. Green, blue, and red ranges show cost ranges in 2100 for SSP1, SSP2, and SSP3, respectively. Global values are represented by solid ranges. Values in the global South are represented by dashed ranges. The diamonds show the costs in the “North America” region. PV – Photovoltaic (Fricko et al., 2016 :cite:`fricko_marker_2016`).
+In :numref:`fig-nonth`, the black ranges show historical cost ranges for 2005. Green, blue, and red ranges show cost ranges in 2100 for SSP1, SSP2, and SSP3, respectively. Global values are represented by solid ranges. Values in the global South are represented by dashed ranges. The diamonds show the costs in the “North America” region. PV – Photovoltaic (Fricko et al., 2017 :cite:`fricko_marker_2017`).
diff --git a/source/energy/conversion/grid.rst b/source/energy/conversion/grid.rst
index 1a294cf..7bb2fa4 100755
--- a/source/energy/conversion/grid.rst
+++ b/source/energy/conversion/grid.rst
@@ -20,7 +20,10 @@ Inter-regional energy transmission infrastructure, such as natural gas pipelines
 
 Systems Integration and Reliability
 ------------------------------------------
-The MESSAGE framework includes a single annual time period characterized by average annual load and 11 geographic regions that span the globe. Seasonal and diurnal load curves and spatial issues such as transmission constraints or renewable resource heterogeneity are treated in a stylized way in the model. The mechanism to represent power system reliability in MESSAGE which is described below, elevates the stylization of temporal resolution by introducing two concepts, peak reserve capacity and general-timescale flexibility, to the model (Sullivan et al., 2013 :cite:`sullivan_electric_2013`). To represent capacity reserves in MESSAGE, a requirement is defined that each region build sufficient firm generating capacity to maintain reliability through reasonable load and contingency events. As a proxy for complex system reliability metrics, a reserve margin-based metric was used, setting the capacity requirement at a multiple of average load, based on electric-system parameters. While many of the same issues apply to both electricity from wind and solar energy, the description below focuses on wind.
+The global MESSAGE model includes a single annual time period within each modeling year characterized by average annual load and 11 geographic regions. 
+Seasonal and diurnal load curves and spatial issues such as transmission constraints or renewable resource heterogeneity are treated in a stylized way in the model. 
+The mechanism to represent power system reliability in MESSAGE is based on (Sullivan et al., 2013 :cite:`sullivan_electric_2013`). This method elevates the stylization of temporal resolution by introducing two concepts, 
+peak reserve capacity and general-timescale flexibility (for mathematical representation see this `Section <https://message.iiasa.ac.at/en/stable/model/MESSAGE/model_core.html#system-reliability-and-flexibility-requirements>`_). To represent capacity reserves in MESSAGE, a requirement is defined that each region build sufficient firm generating capacity to maintain reliability through reasonable load and contingency events. As a proxy for complex system reliability metrics, a reserve margin-based metric was used, setting the capacity requirement at a multiple of average load, based on electric-system parameters. While many of the same issues apply to both electricity from wind and solar energy, the description below focuses on wind.
 
 Toward meeting the firm capacity requirement, conventional generating technologies contribute their nameplate generation capacity while variable renewables contribute a capacity value that declines as the market share of the technology increases. This reflects the fact that wind and solar generators do not always generate when needed, and that their output is generally self-correlated. In order to adjust wind capacity values for different levels of penetration, it was necessary to introduce a stepwise-linear supply curve for wind power (shown in the :numref:`fig-windcap` below). Each bin covers a range of wind penetration levels as fraction of load and has discrete coefficients for the two constraints. The bins are predefined, and therefore are not able to allow, for example, resource diversification to increase capacity value at a given level of wind penetration.
 
diff --git a/source/energy/conversion/other.rst b/source/energy/conversion/other.rst
index 170dcf2..918a2ce 100644
--- a/source/energy/conversion/other.rst
+++ b/source/energy/conversion/other.rst
@@ -9,11 +9,14 @@ Beyond electricity and centralized heat generation there are three further subse
 .. figure:: /_static/costind-other.png
    :width: 700px
 
-   Cost indicators for other conversion technology investment (Fricko et al., 2016 :cite:`fricko_marker_2016`) Abbreviations: CCS – Carbon capture and storage; CTL – Coal to liquids; GTL – Gas to liquids; BTL – Biomass to liquids. 
+   Cost indicators for other conversion technology investment (Fricko et al., 2017 :cite:`fricko_marker_2017`) Abbreviations: CCS – Carbon capture and storage; CTL – Coal to liquids; GTL – Gas to liquids; BTL – Biomass to liquids. 
    
 Liquid Fuel Production
 ----------------------
-Apart from oil refining as predominant supply technology for liquid fuels at present a number of alternative liquid fuel production routes from different feedstocks are represented in MESSAGE (see :numref:`tab-liqfuel`). Different processes for coal liquefaction, gas-to-liquids technologiesand biomass-to-liquids technologies both with and without CCS are covered. Some of these technologies include co-generation of electricity, for example, by burning unconverted syngas from a Fischer-Tropsch synthesis in a gas turbine (c.f. Larson et al., 2012 :cite:`larson_chapter_2012`). Technology costs for the synthetic liquid fuel production options are based on Larson et al. (2012) (:cite:`larson_chapter_2012`).
+Apart from oil refining as predominant supply technology for liquid fuels at present a number of alternative liquid fuel production routes from different feedstocks are represented in MESSAGE 
+(see :numref:`tab-liqfuel`). Different processes for coal liquefaction, gas-to-liquids technologiesand biomass-to-liquids technologies both with and without CCS are covered. 
+Some of these technologies include co-generation of electricity, for example, by burning unconverted syngas from a Fischer-Tropsch synthesis in a gas turbine (c.f. Larson et al., 2012 :cite:`larson_chapter_2012`). 
+Technology costs for the synthetic liquid fuel production options are based on Larson et al. (2012) :cite:`larson_chapter_2012`.
 
 .. _tab-liqfuel:
 .. table:: Liquid fuel production technologies in MESSAGE by energy source.
diff --git a/source/energy/demand.rst b/source/energy/demand.rst
index 9bdea10..d050d53 100755
--- a/source/energy/demand.rst
+++ b/source/energy/demand.rst
@@ -17,11 +17,11 @@ These demands are generated using a so-called scenario generator which is implem
 GDP per capita (PPP) to final energy and, using projections of GDP (PPP) and population, extrapolate the seven energy service demands into the future. The 
 sources for the historical and projected datasets are the following:
 
-1. Historical GDP (PPP) – World Bank (World Development Indicators 2012 :cite:`world_bank_group_world_2012`)
-2. Historical Population – UN Population Division (World Population Projection 2010 :cite:`un_population_division_world_2010`)
-3. Historical Final Energy – International Energy Agency Energy Balances (IEA 2012 :cite:`international_energy_agency_energy_2012`)
-4. Projected GDP (PPP) – Dellink et al (2015 :cite:`dellink_long-term_2015`), see Shared Socio-Economic Pathways database (`SSP scenarios <https://tntcat.iiasa.ac.at/SspDb/>`_)
-5. Projected Population – KC and Lutz (2014 :cite:`kc_human_2014`), see Shared Socio-Economic Pathways database(`SSP scenarios <https://tntcat.iiasa.ac.at/SspDb/>`_)
+1. Historical GDP (PPP) – World Bank (World Development Indicators, 2012 :cite:`world_bank_group_world_2012`)
+2. Historical Population – UN Population Division (World Population Projection, 2010 :cite:`un_population_division_world_2010`)
+3. Historical Final Energy – International Energy Agency Energy Balances (IEA, 2012 :cite:`international_energy_agency_energy_2012`)
+4. Projected GDP (PPP) – Dellink et al. (2015) :cite:`dellink_long-term_2015`, also see Shared Socio-Economic Pathways database (`SSP scenarios <https://tntcat.iiasa.ac.at/SspDb/>`_)
+5. Projected Population – KC and Lutz (2014) :cite:`kc_human_2014`, also see Shared Socio-Economic Pathways database (`SSP scenarios <https://tntcat.iiasa.ac.at/SspDb/>`_)
 
 The scenario generator runs regressions on the historical datasets to establish the relationship for each of the eleven MESSAGE regions between the independent variable (GDP (PPP) per capita) and the following dependent variables:
 
diff --git a/source/energy/enduse/industrial.rst b/source/energy/enduse/industrial.rst
index d8e54c9..f2e6213 100755
--- a/source/energy/enduse/industrial.rst
+++ b/source/energy/enduse/industrial.rst
@@ -15,10 +15,10 @@ While cement production is not explicitly modeled at the process level in MESSAG
 industrial thermal demand in MESSAGEix) and the associated CO2 emissions from the calcination process are accounted for explicitly. In addition, adding carbon capture and storage to
 mitigate these process-based CO2 emission is available.
 
-:numref:`tab-indus` presents the quantitative translation of the the storyline elements of SSP1, SSP2 and SSP3 in terms of electrification rate for industry and feedstocks. These indicators apply to 2010-2100; Intensity improvements are in FE/GDP annually (Fricko et al., 2016 :cite:`fricko_marker_2016`).
+:numref:`tab-indus` presents the quantitative translation of the the storyline elements of SSP1, SSP2 and SSP3 in terms of electrification rate for industry and feedstocks. These indicators apply to 2010-2100; Intensity improvements are in FE/GDP annually (Fricko et al., 2017 :cite:`fricko_marker_2017`).
 
 .. _tab-indus:
-.. table:: Electrification rate within industry and feedstocks for SSP1, SSP2 and SSP3 (Fricko et al., 2016 :cite:`fricko_marker_2016`)
+.. table:: Electrification rate within industry and feedstocks for SSP1, SSP2 and SSP3 (Fricko et al., 2017 :cite:`fricko_marker_2017`)
 
    +--------------------------------+---------------------------------------+-----------------------------------------+---------------------------------------+
    |                                | **SSP1**                              | **SSP2**                                | **SSP3**                              |
diff --git a/source/energy/enduse/resid_commerc.rst b/source/energy/enduse/resid_commerc.rst
index f5f6ee5..5629c99 100755
--- a/source/energy/enduse/resid_commerc.rst
+++ b/source/energy/enduse/resid_commerc.rst
@@ -13,10 +13,10 @@ The residential and commercial thermal energy demand includes fuel switching as
 
 To reflect limitations of switching to alternative fuels, for example as a result of limited infrastructure availability (e.g., district heating network) or some energy carriers being unsuitable for certain applications, share constraints of energy carriers (e.g., electricity) and energy carrier groups (e.g., liquid fuels) are used in the residential and commercial sector. In addition, as in the transport sector, the diffusion speed of alternative fuels is limited to mimic bottlenecks in the supply chains, not explicitly represented in MESSAGEix (e.g., non-energy related infrastructure).
 
-:numref:`tab-rescom` presents the quantitative translation of the the storyline elements of SSP1, SSP2 and SSP3 in terms of electrification rate for the residential and commercial sectors. These indicators apply to 2010-2100; Intensity improvements are in FE/GDP annually (Fricko et al., 2016 :cite:`fricko_marker_2016`).
+:numref:`tab-rescom` presents the quantitative translation of the the storyline elements of SSP1, SSP2 and SSP3 in terms of electrification rate for the residential and commercial sectors. These indicators apply to 2010-2100; Intensity improvements are in FE/GDP annually (Fricko et al., 2017 :cite:`fricko_marker_2017`).
 
 .. _tab-rescom:
-.. table:: Electrification rate within the residential and commercial sectors for SSP1, SSP2 and SSP3 (Fricko et al., 2016 :cite:`fricko_marker_2016`)
+.. table:: Electrification rate within the residential and commercial sectors for SSP1, SSP2 and SSP3 (Fricko et al., 2017 :cite:`fricko_marker_2017`)
 
    +------------------------------+-----------------------------------+------------------------------------+-----------------------------------+
    |                              | **SSP1**                          | **SSP2**                           | **SSP3**                          |
diff --git a/source/energy/enduse/transport.rst b/source/energy/enduse/transport.rst
index 5513733..5cdd047 100755
--- a/source/energy/enduse/transport.rst
+++ b/source/energy/enduse/transport.rst
@@ -15,10 +15,10 @@ Limitations of switching to alternative fuels may occur for example as a result
 
 The demand for international shipping is modeled in a simplified way with a number of different energy carrier options (light and heavy fuel oil, biofuels, natural gas, and hydrogen). The demand for international shipping is coupled to global GDP development with an income elasticity, but to date no demand response in mitigation scenarios is implemented.
 
-:numref:`tab-trans` presents the quantitative translation of the the storyline elements of SSP1, SSP2 and SSP3 in terms of electrification rate for transport (Fricko et al., 2016 :cite:`fricko_marker_2016`).
+:numref:`tab-trans` presents the quantitative translation of the the storyline elements of SSP1, SSP2 and SSP3 in terms of electrification rate for transport (Fricko et al., 2017 :cite:`fricko_marker_2017`).
 
 .. _tab-trans:
-.. table:: Electrification rate within transport for SSP1, SSP2 and SSP3 (Fricko et al., 2016 :cite:`fricko_marker_2016`). The indicators apply to 2010-2100; Intensity improvements are presented in Final Energy (FE)/GDP annually.
+.. table:: Electrification rate within transport for SSP1, SSP2 and SSP3 (Fricko et al., 2017 :cite:`fricko_marker_2017`). The indicators apply to 2010-2100; Intensity improvements are presented in Final Energy (FE)/GDP annually.
 
    +---------------+----------------------------------------+----------------------------------------+---------------------------------------+
    |               | **SSP1**                               | **SSP2**                               | **SSP3**                              |
diff --git a/source/energy/index.rst b/source/energy/index.rst
index c3118bc..fbfbfab 100755
--- a/source/energy/index.rst
+++ b/source/energy/index.rst
@@ -2,9 +2,9 @@
 
 Energy (MESSAGE)
 ==========
-MESSAGE (Model for Energy Supply Strategy Alternatives and their General Environmental Impact) is a linear programming (LP) energy engineering model with global coverage. 
+The `MESSAGEix <https://message.iiasa.ac.at>`_ modeling framework, briefly known as MESSAGE (Model for Energy Supply Strategy Alternatives and their General Environmental Impact), is a linear programming (LP) energy engineering model with global coverage. 
 As a systems engineering optimization model, MESSAGE is primarily used for medium- to long-term energy system planning, energy policy analysis, and scenario development 
-(Messner and Strubegger, 1995 :cite:`messner_users_1995`). The model provides a framework for representing an energy system with all its interdependencies from resource 
+(Huppmann et al., 2019 :cite:`huppmann_message_2019`; Messner and Strubegger, 1995 :cite:`messner_users_1995`). The model provides a framework for representing an energy system with all its interdependencies from resource 
 extraction, imports and exports, conversion, transport, and distribution, to the provision of energy end-use services such as light, space conditioning, industrial 
 production processes, and transportation. In addition, MESSAGE links to GLOBIOM (GLObal BIOsphere Model, cf. Section :ref:`globiom`) to consistently assess the implications 
 of utilizing bioenergy of different types and to integrate the GHG emissions from energy and land use and to the aggregated macro-economic model MACRO (cf. Section :ref:`macro`)
diff --git a/source/energy/resource.rst b/source/energy/resource.rst
index 88c4444..8232890 100755
--- a/source/energy/resource.rst
+++ b/source/energy/resource.rst
@@ -2,7 +2,10 @@ Energy resource endowments
 ==========================
 Fossil Fuel Reserves and Resources
 ---------------------------------------------
-:numref:`tab-globff` shows the assumed total quantities of fossil fuel resources in the MESSAGE model for the base year 2005. :numref:`fig-supply` gives these resource estimates as supply curves. In addition, the assumptions are compared with estimates from the Global Energy Assessment (Rogner et al., 2012 :cite:`rogner_chapter_2012`) as of the year 2009. Estimating fossil fuel reserves is built on both economic and technological assumptions. With an improvement in technology or a change in purchasing power, the amount that may be considered a “reserve” vs. a “resource” (generically referred to here as resources) can actually vary quite widely.
+:numref:`tab-globff` shows the assumed total quantities of fossil fuel resources in the MESSAGE model for the base year 2005. :numref:`fig-supply` gives these resource estimates as supply curves. 
+In addition, the assumptions are compared with estimates from the Global Energy Assessment (Rogner et al., 2012 :cite:`rogner_chapter_2012`) as of the year 2009. Estimating fossil fuel reserves 
+is built on both economic and technological assumptions. With an improvement in technology or a change in purchasing power, the amount that may be considered a “reserve” vs. a “resource” 
+(generically referred to here as resources) can actually vary quite widely.
 
 .. _tab-globff:
 .. list-table:: Assumed global fossil fuel reserves and resources in the MESSAGE model. Estimates from the Global Energy Assessment (Rogner et al., 2012 :cite:`rogner_chapter_2012`) also added for comparison.
diff --git a/source/energy/resource/bioenergy.rst b/source/energy/resource/bioenergy.rst
index 5ed6c0e..d50c33e 100644
--- a/source/energy/resource/bioenergy.rst
+++ b/source/energy/resource/bioenergy.rst
@@ -1,9 +1,14 @@
 Biomass Resources
 ======================
-Biomass energy is another potentially important renewable energy resource in the MESSAGE model. This includes both commercial and non-commercial use. Commercial refers to the use of bioenergy in, for example, power plants or biofuel refineries, while non-commercial refers to the use of bioenergy for residential heating and cooking, primarily in rural households of today’s developing countries. Bioenergy potentials are derived from the GLOBIOM model and differ across SSPs as a result of different levels of competition over land for food and fibre, but ultimately only vary to a limited degree (:numref:`fig-beavail`). The drivers underlying this competition are different land-use developments in the SSPs, which are determined by agricultural productivity and global demand for food consumption. (Fricko et al., 2016 :cite:`fricko_marker_2016`)
+Biomass energy is another potentially important renewable energy resource in the MESSAGE model. This includes both commercial and non-commercial use. 
+Commercial refers to the use of bioenergy in, for example, power plants or biofuel refineries, while non-commercial refers to the use of bioenergy for 
+residential heating and cooking, primarily in rural households of today’s developing countries. Bioenergy potentials are derived from the GLOBIOM model 
+and differ across SSPs as a result of different levels of competition over land for food and fibre, but ultimately only vary to a limited degree (:numref:`fig-beavail`). 
+The drivers underlying this competition are different land-use developments in the SSPs, which are determined by agricultural productivity and global demand for food consumption. (Fricko et al., 2017 :cite:`fricko_marker_2017`)
 
 .. _fig-beavail:
 .. figure:: /_static/Availability_BE.png
    :width: 500px
 
-   Global bioenergy potential. Availability of bioenergy at different price levels in the MESSAGE-GLOBIOM model for the three SSPs (Fricko et al., 2016 :cite:`fricko_marker_2016`). * typically non-commercial biomass is not traded or sold, however in some cases there is a market – prices range from 0.1-1.5$/GJ (Pachauri et al., 2013 :cite:`pachauri_pathways_2013`) ($ equals 2005 USD).
+   Global bioenergy potential. Availability of bioenergy at different price levels in the MESSAGE-GLOBIOM model for the three SSPs (Fricko et al., 2017 :cite:`fricko_marker_2017`). 
+   Typically non-commercial biomass is not traded or sold, however in some cases there is a market – prices range from 0.1-1.5$/GJ (Pachauri et al., 2013 :cite:`pachauri_pathways_2013`) ($ equals 2005 USD).
diff --git a/source/energy/resource/fossilfuel.rst b/source/energy/resource/fossilfuel.rst
index 88cde68..a6bf3aa 100755
--- a/source/energy/resource/fossilfuel.rst
+++ b/source/energy/resource/fossilfuel.rst
@@ -1,25 +1,29 @@
 Fossil Fuel Reserves and Resources
 ====================================
 The availability and costs of fossil fuels influences the future development of the energy system, and therewith future mitigation challenges. Understanding the variations in 
-fossil fuel availability and the underlying extraction cost assumptions across the SSPs is hence important. Our fossil energy resource assumptions in MESSAGE are derived from various sources 
-(Rogner, 1997 :cite:`rogner_assessment_1997`; Riahi et al., 2012 :cite:`riahi_chapter_2012`) and are aligned with the storylines of the individual SSPs. While the physical resource base is identical across the SSPs, considerable differences 
-are assumed regarding the technical and economic availability of overall resources, for example, of unconventional oil and gas. What ultimately determines the attractiveness of a 
-particular type of resource is not just the cost at which it can be brought to the surface, but the cost at which it can be used to provide energy services. Assumptions on fossil 
+fossil fuel availability and the underlying extraction cost assumptions across the SSPs is hence important. Our fossil energy resource assumptions in MESSAGE are derived from various sources, including 
+global databases such as The Federal Institute for Geosciences and Natural Resources (`BGR <https://www.bgr.bund.de/EN/Themen/Energie/energie_node_en.html>`_) and The U.S. Geological Survey 
+(`USGS <https://www.usgs.gov/energy-and-minerals/energy-resources-program/science/energy-resources>`_), as well as market reports and outlooks provided by different energy institutes and agencies. 
+The availability of fossil energy resources in different regions under different socio-economic assumptions are then aligned with the storylines of the individual SSPs 
+(Rogner, 1997 :cite:`rogner_assessment_1997`; Riahi et al., 2012 :cite:`riahi_chapter_2012`). While the physical resource base is identical across the SSPs, considerable differences 
+are assumed regarding the technical and economic availability of overall resources, for example, of unconventional oil and gas. 
+
+What ultimately determines the attractiveness of a particular type of resource is not just the cost at which it can be brought to the surface, but the cost at which it can be used to provide energy services. Assumptions on fossil 
 energy resources should thus be considered together with those on related conversion technologies. In line with the narratives, technological change in fossil fuel extraction and 
 conversion technologies is assumed to be slowest in SSP1, while comparatively faster technological change occurs in SSP3 thereby considerably enlarging the economic potentials of 
-coal and unconventional hydrocarbons (:numref:`tab-globff`, :numref:`fig-supply`). However, driven by tendency toward regional fragmentation the focus in SSP3 is assumed to be on 
+coal and unconventional hydrocarbons (:numref:`tab-globff`, :numref:`fig-supply`). However, driven by the tendency toward regional fragmentation, the focus in SSP3 is assumed to be on 
 developing coal technologies which in the longer term leads to a replacement of oil products by synthetic fuels based on coal-to-liquids technologies. In contrast, for SSP2 we 
 assume a continuation of recent trends, focusing more on developing extraction technologies for unconventional hydrocarbon resources, thereby leading to higher potential cumulative oil 
-extraction than in the other SSPs (:numref:`fig-supply`, middle panel).
+extraction than in the other SSPs (:numref:`fig-supply`, the middle panel).
 
-:numref:`tab-globff` shows the assumed total quantities of fossil fuel resources in the MESSAGE model for the base year 2005. :numref:`fig-supply` gives these resource estimates as cumulative resource supply 
-curves. In addition, the assumptions are compared with estimates from the Global Energy Assessment (Rogner et al., 2012 :cite:`rogner_chapter_2012`) as of the year 2009. Estimating 
+:numref:`tab-globff` shows the assumed total quantities of fossil fuel resources in the MESSAGE model for 2005. :numref:`fig-supply` gives these resource estimates as cumulative resource supply 
+curves. In addition, the assumptions are compared with estimates from the Global Energy Assessment (Rogner et al., 2012 :cite:`rogner_chapter_2012`) and the databases mentioned earlier. Estimating 
 fossil fuel reserves is built on both economic and technological assumptions. With an improvement in technology or a change in purchasing power, the amount that may be considered a 
 “reserve” vs. a “resource” (generically referred to here as resources) can actually vary quite widely.
 
 ‘Reserves’ are generally defined as being those quantities for which geological and 
 engineering information indicate with reasonable certainty that they can be recovered in the future from known reservoirs under existing economic and operating conditions. 
-‘Resources’ are detected quantities that cannot be profitably recovered with current technology, but might be recoverable in the future, as well as those quantities that are geologically 
+‘Resources’ are detected quantities that cannot be profitably recovered with the current technology, but might be recoverable in the future, as well as those quantities that are geologically 
 possible, but yet to be found. The remainder are ‘Undiscovered resources’ and, by definition, one can only speculate on their existence. Definitions are based on Rogner et al. (2012) 
 :cite:`rogner_chapter_2012`. 
 
@@ -60,7 +64,7 @@ possible, but yet to be found. The remainder are ‘Undiscovered resources’ an
 The following table (:numref:`tab-ffavail`) presents the ultimate fossil resource availability for coal, oil and gas, for SSP1, SSP2 and SSP3, respectively.
 
 .. _tab-ffavail:
-.. list-table:: Fossil resource availability for SSP1, SSP2, and SSP3 (Fricko et al., 2016 :cite:`fricko_marker_2016`).
+.. list-table:: Fossil resource availability for SSP1, SSP2, and SSP3 (Fricko et al., 2017 :cite:`fricko_marker_2017`).
    :widths: 20 20 20 20
    :header-rows: 1
 
@@ -82,22 +86,22 @@ The following table (:numref:`tab-ffavail`) presents the ultimate fossil resourc
      - 24
 
 Coal is the largest resource among fossil fuels; it accounts for more than 50% of total fossil reserve plus resource estimates even at the higher end of the assumptions, which includes 
-considerable amounts of unconventional hydrocarbons. Oil is the most vulnerable fossil fuel at less than 10 ZJ of conventional oil and possibly less than 10 ZJ of unconventional oil. 
+considerable amounts of unconventional hydrocarbons. Oil is the fastest depleting fossil fuel with less than 10 ZJ of conventional oil and possibly less than 10 ZJ of unconventional oil. 
 Natural gas is more abundant in both the conventional and unconventional categories.
 
 :numref:`fig-supply` presents the cumulative global resource supply curves for coal, oil and gas in the IIASA IAM framework. Green shaded resources are technically and economically 
 extractable in all SSPs, purple shaded resources are additionally available in SSP1 and SSP2 and blue shaded resources are additionally available in SSP2. Coloured vertical lines 
-represent the cumulative use of each resource between 2010 and 2100 in the SSP baselines (see top panel for colour coding), and are thus the result of the combined effect of the
+represent the cumulative use of each resource between 2010 and 2100 in the SSP baselines (see the top panel for colour coding), and are thus the result of the combined effect of the
 assumptions on fossil resource availability and conversion technologies in the SSP baseline scenarios. 
 
 .. _fig-supply:
 .. figure:: /_static/GlobalResourceSupplyCurves.png
    :width: 750px
    
-   Cumulative global resource supply curves for coal (top), oil (middle), and gas (bottom) in the IIASA IAM framework (Fricko et al., 2016 :cite:`fricko_marker_2016`). 
+   Cumulative global resource supply curves for coal (top), oil (middle), and gas (bottom) in the IIASA IAM framework (Fricko et al., 2017 :cite:`fricko_marker_2017`). 
    
 Conventional oil and gas are distributed unevenly throughout the world, with only a few regions dominating the reserves. Nearly half of the reserves of conventional oil is found in 
-Middle East and North Africa, and close to 40% of conventional gas is found in Russia and the former Soviet Union states. The situation is somewhat different for unconventional oil 
+Middle East and North Africa, and close to 40% of conventional gas is found in Russia and the Former Soviet Union states. The situation is somewhat different for unconventional oil 
 of which North and Latin America potentially possess significantly higher global shares. Unconventional gas in turn is distributed quite evenly throughout the world, with North America 
 holding most (roughly 25% of global resources). The distribution of coal reserves shows the highest geographical diversity which in the more fragmented SSP3 world contributes to 
 increased overall reliance on this resource. Russia and the former Soviet Union states, Pacific OECD, North America, and Centrally Planned Asia and China all possess more than 10 ZJ 
diff --git a/source/energy/resource/nuclear.rst b/source/energy/resource/nuclear.rst
index dbc8e67..edfef23 100755
--- a/source/energy/resource/nuclear.rst
+++ b/source/energy/resource/nuclear.rst
@@ -14,6 +14,6 @@ and fuel cycle are modeled at the global level.
 
    Global uranium resources in the MESSAGE interpretation of the SSPs compared to seven supply curves from a literature review (Schneider and Sailor, 2008 :cite:`schneider_long-term_2008`). 
    Conservative Crustal and Optimistic Crustal refer to simple crustal models of uranium distribution in the crust and the of extraction costs on the concentration. Pure-KCR refers to a fit 
-   of a simple crustal model to known conventional resources (KCR) as estimated by the Red Book 2003 (OECD/NEA 2004, :cite:`oecd_uranium_2004`). PPM-Cost over the simple crustal models 
+   of a simple crustal model to known conventional resources (KCR) as estimated by the Red Book 2003 (OECD/NEA, 2004 :cite:`oecd_uranium_2004`). PPM-Cost over the simple crustal models 
    include a relationship between uranium grade and extraction costs. FCCCG(1) and (2) as well as DANESS refer to estimats from more complicated models of the dependency of extraction costs on
    uranium concentration (and therefore resource grade).
diff --git a/source/energy/resource/renewable.rst b/source/energy/resource/renewable.rst
index fca42d2..a9f121f 100755
--- a/source/energy/resource/renewable.rst
+++ b/source/energy/resource/renewable.rst
@@ -2,11 +2,11 @@
 
 Non-Biomass Renewable Resources
 ================================
-:numref:`tab-depl` shows the assumed total potentials of non-biomass renewable energy deployment (by resource type) in the MESSAGE model. In addition, the assumptions are compared 
-with technical potential estimates from the Global Energy Assessment (Rogner et al., 2012 :cite:`rogner_chapter_2012`). In this context, it is important to note that typical MESSAGE 
+:numref:`tab-depl` shows the assumed total potentials of non-biomass renewable energy deployment (by resource type) in the MESSAGE model. In addition, the technical potential estimates are based on different sources, 
+such as the U.S. National Renewable Energy Laboratory `database <https://www.nrel.gov/grid/data-tools.html>`_ as described in the Global Energy Assessment (Rogner et al., 2012 :cite:`rogner_chapter_2012`). 
+In this context, it is important to note that typical MESSAGE 
 scenarios do not consider the full technical potential of renewable energy resources, but rather only a subset of those potentials, owing to additional constraints (e.g., sustainability 
-criteria, technology diffusion and systems integration issues, and other economic considerations) that may not be fully captured within the model. These constraints may lead to a 
-significant reduction of the technical potential.
+criteria, technology diffusion and systems integration issues, and other economic considerations). These constraints may lead to a significant reduction of the technical potential.
 
 .. _tab-depl:
 .. list-table:: Assumed global non-biomass renewable energy deployment potentials in the MESSAGE model. Estimates from the Global Energy Assessment (Rogner et al., 2012  :cite:`rogner_chapter_2012`) also added for comparison.
@@ -35,10 +35,14 @@ significant reduction of the technical potential.
      - 23
      - 810 - 1400
 
-*Notes: MESSAGE renewable energy potentials are based on* Pietzcker et al. (2014) :cite:`pietzcker_solar_2014`, Eurek et al. (in review) :cite:`eurek_wind_2016`, Christiansson (1995) :cite:`christiansson_diffusion_1995`, *and* Rogner et al (2012) :cite:`rogner_chapter_2012`. *The potentials for non-combustible renewable energy sources are specified in terms of the electricity or heat that can be produced by specific technologies (i.e., from a secondary energy perspective). By contrast, the technical potentials from* :cite:`rogner_chapter_2012` *refer to the flows of energy that could become available as inputs for technology conversion. So for example, the technical potential for wind is given as the kinetic energy available for wind power generation, whereas the deployment potential would only be the electricity that could be generated by the wind turbines.*
+*Notes: MESSAGE renewable energy potentials are estimated based on the methods explained in* Pietzcker et al., 2014 :cite:`pietzcker_solar_2014`, Eurek et al., 2017 :cite:`eurek_wind_2017`, 
+Christiansson, 1995 :cite:`christiansson_diffusion_1995`, *and* Rogner et al., 2012 :cite:`rogner_chapter_2012`. *The potentials for non-combustible renewable energy sources are specified 
+in terms of the electricity or heat that can be produced by specific technologies (i.e., from a secondary energy perspective). By contrast, the technical potentials from* :cite:`rogner_chapter_2012` 
+*refer to the flows of energy that could become available as inputs for technology conversion. So for example, the technical potential for wind is given as the kinetic energy available for wind power 
+generation, whereas the deployment potential would only be the electricity that could be generated by the wind turbines.*
 
 Regional resource potentials for solar and wind are classified according to resource quality (annual capacity factor) based on Pietzcker et al. (2014, :cite:`pietzcker_solar_2014`) and 
-Eurek et al. (in review, :cite:`eurek_wind_2016`). Regional resource potentials as implemented into MESSAGE are provided by region and capacity factor for solar PV, concentrating solar 
+Eurek et al. (2017) :cite:`eurek_wind_2017`. Regional resource potentials as implemented into MESSAGE are provided by region and capacity factor for solar PV, concentrating solar 
 power (CSP), and onshore/offshore wind in Johnson et al. (2016, :cite:`johnson_vre_2016`). The physical potential of these sources is assumed to be the same across all SSPs.  
 :numref:`tab-pv`, :numref:`tab-csp`, :numref:`tab-onshorewind`, :numref:`tab-offshorewind` show the resource potential for solar PV, CSP (solar multiples (SM) of 1 & 3), on- and offshore wind respectivey. For wind, 
 :numref:`tab-capfactonshore` and :numref:`tab-capfactoffshore` list the capacity factors corresponding to the wind classes used in the resource tables. It is important to note that part of 
@@ -46,7 +50,7 @@ the resource that is useable at economically competitive costs is assumed to dif
 
 
 .. _tab-pv:
-.. table:: Resource potential (EJ) by region and capacity factor for solar photovoltaic (PV) technology (Johnson et al. 2016, :cite:`johnson_vre_2016`). For a description of each of the regions represented in the table, see :ref:`spatial`.
+.. table:: Resource potential (EJ) by region and capacity factor for solar photovoltaic (PV) technology (Johnson et al., 2016 :cite:`johnson_vre_2016`). For a description of each of the regions represented in the table, see :ref:`spatial`.
 
    +-----------+----------+---------------------------------------------------------------------------------+
    |                      |                                                                                 |
@@ -109,7 +113,7 @@ the resource that is useable at economically competitive costs is assumed to dif
 
 
 .. _tab-csp:
-.. table:: Resource potential (EJ) by region and capacity factor for concentrating solar power (CSP) technologies with solar multiples (SM) of 1 and 3  (Johnson et al. 2016, :cite:`johnson_vre_2016`).
+.. table:: Resource potential (EJ) by region and capacity factor for concentrating solar power (CSP) technologies with solar multiples (SM) of 1 and 3  (Johnson et al., 2016 :cite:`johnson_vre_2016`).
 
    +--------------+----------+---------------------------------------------------------------------------------------+
    |                         |                                                                                       |
@@ -176,7 +180,7 @@ the resource that is useable at economically competitive costs is assumed to dif
 
 
 .. _tab-onshorewind:
-.. table:: Resource potential (EJ) by region and wind class for onshore wind (Johnson et al. 2016, :cite:`johnson_vre_2016`).
+.. table:: Resource potential (EJ) by region and wind class for onshore wind (Johnson et al., 2016 :cite:`johnson_vre_2016`).
 
    +---------+-------------------------------------------------------+
    |         |                                                       |
@@ -239,7 +243,7 @@ the resource that is useable at economically competitive costs is assumed to dif
 
 
 .. _tab-capfactonshore:
-.. table:: Capacity factor by region and wind class for onshore wind (Johnson et al. 2016, :cite:`johnson_vre_2016`).
+.. table:: Capacity factor by region and wind class for onshore wind (Johnson et al., 2016 :cite:`johnson_vre_2016`).
 
    +-------+-----------------------------------------------------+
    |       |                                                     |
@@ -298,7 +302,7 @@ the resource that is useable at economically competitive costs is assumed to dif
 
 
 .. _tab-offshorewind:
-.. table:: Resource potential (EJ) by region and wind class for offshore wind (Johnson et al. 2016, :cite:`johnson_vre_2016`).
+.. table:: Resource potential (EJ) by region and wind class for offshore wind (Johnson et al., 2016 :cite:`johnson_vre_2016`).
 
    +---------+-----------------------------------------------------+
    |         |                                                     |
@@ -361,7 +365,7 @@ the resource that is useable at economically competitive costs is assumed to dif
 
 
 .. _tab-capfactoffshore:
-.. table:: Capacity factor by region and wind class for offshore wind (Johnson et al. 2016, :cite:`johnson_vre_2016`).
+.. table:: Capacity factor by region and wind class for offshore wind (Johnson et al., 2016 :cite:`johnson_vre_2016`).
 
    +---------+-----------------------------------------------------+
    |         |                                                     |
diff --git a/source/energy/tech.rst b/source/energy/tech.rst
index cda7184..4ef7bf4 100755
--- a/source/energy/tech.rst
+++ b/source/energy/tech.rst
@@ -23,7 +23,7 @@ activity in the previous period *t-1* (Messner and Strubegger, 1995 :cite:`messn
 
 While limiting the possibility of flip-flop behavior as is frequently observed in unconstrained Linear Programming (LP) models such as MESSAGE, a drawback of such hard 
 growth constraints is that the relative advantage of some technology over another technology is not taken into account and therefore even for very competitive technologies, 
-no acceleration of technology diffusion is possible. In response to this limitation, so called flexible or soft dynamic constraints have been introduced into MESSAGE 
+no rapid acceleration of technology diffusion is possible. In response to this limitation, so called flexible or soft dynamic constraints have been introduced into MESSAGE 
 (Keppo and Strubegger, 2010 :cite:`keppo_short_2010`). These allow faster technology diffusion at additional costs and therefore generate additional model flexibility 
 while still reducing the flip-flop behavior and sudden penetration of technologies.
 
diff --git a/source/land_use/crop.rst b/source/land_use/crop.rst
index 9b9c219..b0e36e6 100755
--- a/source/land_use/crop.rst
+++ b/source/land_use/crop.rst
@@ -3,4 +3,16 @@
 Crop production
 ----
 
-GLOBIOM directly represents production from three major land cover types: cropland, managed forest, and areas suitable for short rotation tree plantations. Crop production accounts for more than 30 of the globally most important crops. The average yield level for each crop in each country is taken from FAOSTAT. Management related yield coefficients according to fertilizer and irrigation rates are explicitly simulated with the EPIC model (Williams and Singh 1995 :cite:`williams_computer_1995`) for 17 crops (barley, dry beans, cassava, chickpea, corn, cotton, ground nuts, millet, potatoes, rapeseed, rice, soybeans, sorghum, sugarcane, sunflower, sweet potatoes, and wheat). These 17 crops together represent nearly 80 % of the 2007 harvested area and 85% of the vegetal calorie supply as reported by FAOSTAT. Four management systems are considered (irrigated, high input - rainfed, low input - rainfed and subsistence management systems) corresponding to the International Food and Policy Research Institute (IFPRI) crop distribution data classification (You and Wood, 2006 :cite:`you_entropy_2006`). Within each management system, input structure is fixed following a Leontieff production function. But crop yields can change in reaction to external socio-economic drivers through switch to another management system or reallocation of the production to a more or less productive Supply Unit. Besides the endogennous mechanisms, an exogenous component representing long-term technological change is also considered. Only two management systems are differentiated for the remaining crops (bananas, other dry beans, coconuts, coffee, lentils, mustard seed, olives, oil palm, plantains, peas, other pulses, sesame seed, sugar beet, and yams) – rainfed and irrigated. Rainfed and irrigated crop yield coefficients, and crop specific irrigation water requirements for crops not simulated with EPIC, and costs for four irrigation systems for all crops, are derived from a variety of sources as described in Sauer et al. (2008 :cite:`sauer_agriculture_2008`). Crop supply can enter one of three processing/demand channels: consumption, livestock production and biofuel production (see :numref:`fig-landuse_product_structure`). 
+GLOBIOM directly represents production from three major land cover types: cropland, managed forest, and areas suitable for short rotation tree plantations. Crop production accounts for more than 30 of the 
+globally most important crops. The average yield level for each crop in each country is taken from FAOSTAT. Management related yield coefficients according to fertilizer and irrigation rates are explicitly 
+simulated with the EPIC model (Williams and Singh, 1995 :cite:`williams_computer_1995`) for 17 crops (barley, dry beans, cassava, chickpea, corn, cotton, ground nuts, millet, potatoes, rapeseed, rice, soybeans, 
+sorghum, sugarcane, sunflower, sweet potatoes, and wheat). These 17 crops together represent nearly 80 % of the 2007 harvested area and 85% of the vegetal calorie supply as reported by FAOSTAT. Four management 
+systems are considered (irrigated, high input - rainfed, low input - rainfed and subsistence management systems) corresponding to the International Food and Policy Research Institute (IFPRI) crop distribution data 
+classification (You and Wood, 2006 :cite:`you_entropy_2006`). Within each management system, input structure is fixed following a Leontieff production function. But crop yields can change in reaction to external 
+socio-economic drivers through switch to another management system or reallocation of the production to a more or less productive Supply Unit. 
+
+Besides the endogennous mechanisms, an exogenous component representing 
+long-term technological change is also considered. Only two management systems are differentiated for the remaining crops (bananas, other dry beans, coconuts, coffee, lentils, mustard seed, olives, oil palm, plantains, 
+peas, other pulses, sesame seed, sugar beet, and yams) – rainfed and irrigated. Rainfed and irrigated crop yield coefficients, and crop specific irrigation water requirements for crops not simulated with EPIC, 
+and costs for four irrigation systems for all crops, are derived from a variety of sources as described in Sauer et al. (2008) :cite:`sauer_agriculture_2008`. Crop supply can enter one of three processing/demand 
+channels: consumption, livestock production and biofuel production (see :numref:`fig-landuse_product_structure`). 
diff --git a/source/land_use/food.rst b/source/land_use/food.rst
index 14a1915..741e768 100755
--- a/source/land_use/food.rst
+++ b/source/land_use/food.rst
@@ -2,7 +2,15 @@
 
 Food demand
 -----------
-Food demand is in GLOBIOM endogenous and depends on population, gross domestic product (GDP) and own produt price. Population and GDP are exogenous variables while prices are endogenous. The simple demand system is presented in Eq. :eq:`foodelasticity`. First, for each product :math:`i` in region :math:`r` and period :math:`t`,  the prior demand quantity :math:`Q` is calculated as a function of population POP, GDP per capita :math:`GDP^{cap}` adjusted by the income elasticity :math:`\varepsilon^{GDP}`, and the base year consumption level as reported in the Food Balance Sheets of FAOSTAT. If the prior demand quantity could be satisfied at the base year price :math:`P`, this would be also the optimal demand quantity :math:`Q`. However, usually the optimal quantity will be different from the prior quantity, and will depend on the optimal price :math:`P` and the price elasticity :math:`\varepsilon^{price}`, the latter calculated from USDA (Seale, Regmi et al. 2003, :cite:`seale_international_2003`), updated in Muhammad, Seale et al. (2011, :cite:`muhammad_international_2011`) for the base year 2000. Because food demand in developed countries is more inelastic than in developing ones, the value of this elasticity is assumed to decrease with the level of GDP per capita. The rule applied is that the price elasticity of developing countries converges to the price elasticity of the USA in 2000 at the same pace as their GDP per capita reach the USA GDP per capita value of 2000. This allows capturing the effect of change in relative prices on food consumption taking into account heterogeneity of responses across regions, products and over time.
+Food demand is in GLOBIOM endogenous and depends on population, gross domestic product (GDP) and own produt price. Population and GDP are exogenous variables while prices are endogenous. 
+The simple demand system is presented in Eq. :eq:`foodelasticity`. First, for each product :math:`i` in region :math:`r` and period :math:`t`,  the prior demand quantity :math:`Q` is calculated as a 
+function of population POP, GDP per capita :math:`GDP^{cap}` adjusted by the income elasticity :math:`\varepsilon^{GDP}`, and the base year consumption level as reported in the Food Balance Sheets of FAOSTAT. 
+If the prior demand quantity could be satisfied at the base year price :math:`P`, this would be also the optimal demand quantity :math:`Q`. However, usually the optimal quantity will be different from the prior 
+quantity, and will depend on the optimal price :math:`P` and the price elasticity :math:`\varepsilon^{price}`, the latter calculated from USDA (Seale et al., 2003 :cite:`seale_international_2003`), 
+updated in Muhammad et al. (2011) :cite:`muhammad_international_2011` for the base year 2000. Because food demand in developed countries is more inelastic than in developing ones, 
+the value of this elasticity is assumed to decrease with the level of GDP per capita. The rule applied is that the price elasticity of developing countries converges to the price elasticity of the USA in 
+2000 at the same pace as their GDP per capita reach the USA GDP per capita value of 2000. This allows capturing the effect of change in relative prices on food consumption taking into account heterogeneity 
+of responses across regions, products and over time.
 
 .. math:: \frac{Q_{i,r,t}}{\overline{Q}_{i,r,t}} = \left( \frac{P_{i,r,t}}{\overline{P}_{i,r,2000}} \right)^{\varepsilon_{i,r,t}^{price}}
    :label: foodelasticity
diff --git a/source/land_use/forest.rst b/source/land_use/forest.rst
index 2d660a7..a9b0916 100755
--- a/source/land_use/forest.rst
+++ b/source/land_use/forest.rst
@@ -3,9 +3,27 @@
 Forestry
 --------
 
-The forestry sector is represented in GLOBIOM with five categories of primary products (pulp logs, saw logs, biomass for energy, traditional fuel wood, and other industrial logs) which are consumed by industrial energy, cooking fuel demand, or processed and sold on the market as final products (wood pulp and sawnwood). These products are supplied from managed forests and short rotation plantations. Harvesting cost and mean annual increments are informed by the G4M global forestry model (Kindermann, Obersteiner et al. 2006 :cite:`kindermann_predicting_2006`) which in turn calculates them based on thinning strategies and length of the rotation period.
+The forestry sector is represented in GLOBIOM with five categories of primary products (pulp logs, saw logs, biomass for energy, traditional fuel wood, and other industrial logs) which are consumed by industrial energy, 
+cooking fuel demand, or processed and sold on the market as final products (wood pulp and sawnwood). These products are supplied from managed forests and short rotation plantations. Harvesting cost and mean annual 
+increments are informed by the G4M global forestry model (Kindermann et al., 2006 :cite:`kindermann_predicting_2006`) which in turn calculates them based on thinning strategies and length of the rotation period.
 
-Primary forest production from traditional managed forests is characterized also at the level of SimUs. The most important parameters for the model are mean annual increment, maximum share of saw logs in the mean annual increment, and harvesting cost. These parameters are shared with the G4M model – a successor of the model described by Kindermann et al. (2006 :cite:`kindermann_predicting_2006`). More specifically, mean annual increment for the current management, is obtained by downscaling biomass stock data from the Global Forest Resources Assessment (FAO, 2006 :cite:`FAO_global_2006`) from the country level to a 0.5 x 0.5 degree grid using the method described in Kindermann et al. (2008 :cite:`kindermann_global_forest_2008`). The downscaled biomass stock data is subsequently used to parameterize increment curves. Finally, the saw logs share is estimated by the tree size, which in turn depends on yield and rotation time. Harvesting costs are adjusted for slope and tree size as well. 
-Among the five primary forest products, saw logs, pulp logs and biomass for energy are further processed. Sawn wood and wood pulp production and demand parameters rely on the 4DSM model described in Rametsteiner et al. (2007 :cite:`rametsteiner_study_2007`). FAO data and other secondary sources have been used for quantities and prices of sawn wood and wood pulp. For processing cost estimates of these products an internal IIASA database and proprietary data (e.g. RISI database for locations of individual pulp and paper mills, with additional economic and technical information, http://www.risiinfo.com) were used. Biomass for energy can be converted in several processes: combined heat and power production, fermentation for ethanol, heat, power and gas production, and gasification for methanol and heat production. Processing cost and conversion coefficients are obtained from various sources (Biomass Technology Group, 2005 :cite:`biomass_handbook_2005`; Hamelinck and Faaij, 2001 :cite:`hamelinck_future_2001`; Leduc et al., 2008 :cite:`leduc_optimal_2008`; Sorensen, 2005 :cite:`sorensen_economies_2005`). Demand for woody bioenergy production is implemented through minimum quantity constraints, similar to demand for other industrial logs and for firewood.
-Woody biomass for bioenergy can also be produced on short rotation tree plantations. To parameterize this land use type in terms of yields, an evaluation of the land availability and suitability was carried out. Calculated plantation costs involve the establishment cost and the harvesting cost. The establishment related capital cost includes only sapling cost for manual planting (Carpentieri et al., 1993 :cite:`carpentieri_future_1993`; Herzogbaum GmbH, 2008 :cite:`herzogbaum_forstpflanzen_2008`). Labour requirements for plantation establishment are based on Jurvelius (1997 :cite:`jurvelius_labor_1997`), and consider land preparation, saplings transport, planting and fertilization. These labour requirements are adjusted for temperate and boreal regions to take into account the different site conditions. The average wages for planting are obtained from ILO (2007 :cite:`ILO_occupational_2007`). 
-Harvesting cost includes logging and timber extraction. The unit cost of harvesting equipment and labour is derived from various datasets for Europe and North America (e.g. FPP, 1999 :cite:`FPP_holzernte_1999`; Jiroušek et al., 2007 :cite:`jiroušek_productivity_2007` ; Stokes et al., 1986 :cite:`stokes_field_1986` ; Wang et al., 2004 :cite:`wang_productivity_2004`). Because the productivity of harvesting equipment depends on terrain conditions, a slope factor (Hartsough et al., 2001 :cite:`hartsough_harvesting_2001`) was integrated to estimate total harvesting cost. The labour cost, as well as the cost of saplings, is regionally adjusted by the ratio of mean PPP (purchasing power parity over GDP), (Heston et al., 2006 :cite:`heston_penn_2006`).
+Primary forest production from traditional managed forests is characterized also at the level of SimUs. The most important parameters for the model are mean annual increment, maximum share of saw logs in the mean annual 
+increment, and harvesting cost. These parameters are shared with the G4M model – a successor of the model described by Kindermann et al. (2006) :cite:`kindermann_predicting_2006`. More specifically, mean annual increment 
+for the current management, is obtained by downscaling biomass stock data from the Global Forest Resources Assessment (FAO, 2006 :cite:`FAO_global_2006`) from the country level to a 0.5 x 0.5 degree grid using the method 
+described in Kindermann et al. (2008) :cite:`kindermann_global_forest_2008`. The downscaled biomass stock data is subsequently used to parameterize increment curves. Finally, the saw logs share is estimated by the tree size, 
+which in turn depends on yield and rotation time. Harvesting costs are adjusted for slope and tree size as well. 
+Among the five primary forest products, saw logs, pulp logs and biomass for energy are further processed. Sawn wood and wood pulp production and demand parameters rely on the 4DSM model described in 
+Rametsteiner et al. (2007) :cite:`rametsteiner_study_2007`. FAO data and other secondary sources have been used for quantities and prices of sawn wood and wood pulp. For processing cost estimates of these products an internal 
+IIASA database and proprietary data (e.g. RISI database for locations of individual pulp and paper mills, with additional economic and technical information, http://www.risiinfo.com) were used. Biomass for energy can be converted 
+in several processes: combined heat and power production, fermentation for ethanol, heat, power and gas production, and gasification for methanol and heat production. Processing cost and conversion coefficients are obtained from 
+various sources (Biomass Technology Group, 2005 :cite:`biomass_handbook_2005`; Hamelinck and Faaij, 2001 :cite:`hamelinck_future_2001`; Leduc et al., 2008 :cite:`leduc_optimal_2008`; Sorensen, 2005 :cite:`sorensen_economies_2005`). 
+Demand for woody bioenergy production is implemented through minimum quantity constraints, similar to demand for other industrial logs and for firewood.
+Woody biomass for bioenergy can also be produced on short rotation tree plantations. To parameterize this land use type in terms of yields, an evaluation of the land availability and suitability was carried out. 
+Calculated plantation costs involve the establishment cost and the harvesting cost. The establishment related capital cost includes only sapling cost for manual planting 
+(Carpentieri et al., 1993 :cite:`carpentieri_future_1993`; Herzogbaum GmbH, 2008 :cite:`herzogbaum_forstpflanzen_2008`). Labour requirements for plantation establishment are based on Jurvelius (1997) :cite:`jurvelius_labor_1997`, 
+and consider land preparation, saplings transport, planting and fertilization. These labour requirements are adjusted for temperate and boreal regions to take into account the different site conditions. 
+The average wages for planting are obtained from ILO (2007) :cite:`ILO_occupational_2007`. 
+Harvesting cost includes logging and timber extraction. The unit cost of harvesting equipment and labour is derived from various datasets for Europe and North America 
+(e.g. FPP, 1999 :cite:`FPP_holzernte_1999`; Jiroušek et al., 2007 :cite:`jiroušek_productivity_2007` ; Stokes et al., 1986 :cite:`stokes_field_1986` ; Wang et al., 2004 :cite:`wang_productivity_2004`). 
+Because the productivity of harvesting equipment depends on terrain conditions, a slope factor (Hartsough et al., 2001 :cite:`hartsough_harvesting_2001`) was integrated to estimate total harvesting cost. 
+The labour cost, as well as the cost of saplings, is regionally adjusted by the ratio of mean PPP (purchasing power parity over GDP), (Heston et al., 2006 :cite:`heston_penn_2006`).
diff --git a/source/land_use/index.rst b/source/land_use/index.rst
index fb5e35e..cc7aa37 100755
--- a/source/land_use/index.rst
+++ b/source/land_use/index.rst
@@ -2,9 +2,20 @@
 
 Land-use (GLOBIOM)
 ==================
-Land-use dynamics are modelled with the GLOBIOM (GLobal BIOsphere Management) model, which is a partial-equilibrium model (Havlik et al., 2011 :cite:`havlik_global_2011`; Havlik et al., 2014 :cite:`havlik_climate_2014`). GLOBIOM represents the competition between different land-use based activities. It includes a detailed representation of the agricultural, forestry and bio-energy sector, which allows for the inclusion of detailed grid-cell information on biophysical constraints and technological costs, as well as a rich set of environmental parameters, incl. comprehensive AFOLU (agriculture, forestry and other land use) GHG emission accounts and irrigation water use. For spatially explicit projections of the change in afforestation, deforestation, forest management, and their related CO2 emissions, GLOBIOM is coupled with the G4M (Global FORest Model) model (Kindermann et al., 2006 :cite:`kindermann_predicting_2006`; Kindermann et al., 2008 :cite:`kindermann_global_2008`; Gusti, 2010 :cite:`gusti_algorithm_2010`). The spatially explicit G4M model compares the income of forest (difference of wood price and harvesting costs, income by storing carbon in forests) with income by alternative land use on the same place, and decides on afforestation, deforestation or alternative management options. As outputs, G4M provides estimates of forest area change, carbon uptake and release by forests, and supply of biomass for bioenergy and timber.
-
-As a partial equilibrium model representing land-use based activities, including agriculture, forestry and bioenergy sectors (see :numref:`fig-landuse_product_structure`), production adjusts to meet the demand at the level of 30 economic regions (see list of the regions in Section :ref:`spatial`). International trade representation is based on the spatial equilibrium modelling approach, where individual regions trade with each other based purely on cost competitiveness because goods are assumed to be homogenous (Takayama and Judge 1971 :cite:`takayama_spatial_1971`; Schneider, McCarl et al. 2007 :cite:`schneider_agricultural_2007`). Market equilibrium is determined through mathematical optimization which allocates land and other resources to maximize the sum of consumer and producer surplus (McCarl and Spreen 1980 :cite:`mccarl_surplus_1980`). As in other partial equilibrium models, prices are endogenous. The model is run recursively dynamic with a 10 year time step, going from 2000 to 2100. The model is solved using a linear programming solver and can be run on a personal computer with the GAMS software.
+Land-use dynamics are modelled with the GLOBIOM (GLobal BIOsphere Management) model, which is a partial-equilibrium model (Havlik et al., 2011 :cite:`havlik_global_2011`; Havlik et al., 2014 :cite:`havlik_climate_2014`). 
+GLOBIOM represents the competition between different land-use based activities. It includes a detailed representation of the agricultural, forestry and bio-energy sector, which allows for the inclusion of detailed grid-cell 
+information on biophysical constraints and technological costs, as well as a rich set of environmental parameters, incl. comprehensive AFOLU (agriculture, forestry and other land use) GHG emission accounts and irrigation water use. 
+For spatially explicit projections of the change in afforestation, deforestation, forest management, and their related CO2 emissions, GLOBIOM is coupled with the G4M (Global FORest Model) model (Kindermann et al., 2006 :cite:`kindermann_predicting_2006`; 
+Kindermann et al., 2008 :cite:`kindermann_global_2008`; Gusti, 2010 :cite:`gusti_algorithm_2010`). The spatially explicit G4M model compares the income of forest (difference of wood price and harvesting costs, income by 
+storing carbon in forests) with income by alternative land use on the same place, and decides on afforestation, deforestation or alternative management options. 
+As outputs, G4M provides estimates of forest area change, carbon uptake and release by forests, and supply of biomass for bioenergy and timber.
+
+As a partial equilibrium model representing land-use based activities, including agriculture, forestry and bioenergy sectors (see :numref:`fig-landuse_product_structure`), 
+production adjusts to meet the demand at the level of 30 economic regions (see list of the regions in Section :ref:`spatial`). International trade representation is based on 
+the spatial equilibrium modelling approach, where individual regions trade with each other based purely on cost competitiveness because goods are assumed to be homogenous 
+(Takayama and Judge, 1971 :cite:`takayama_spatial_1971`; Schneider, McCarl et al., 2007 :cite:`schneider_agricultural_2007`). Market equilibrium is determined through mathematical optimization which allocates land 
+and other resources to maximize the sum of consumer and producer surplus (McCarl and Spreen, 1980 :cite:`mccarl_surplus_1980`). As in other partial equilibrium models, prices are endogenous. 
+The model is run recursively dynamic with a 10 year time step, going from 2000 to 2100. The model is solved using a linear programming solver and can be run on a personal computer with the GAMS software.
 
 .. _fig-landuse_product_structure:
 .. figure:: /_static/GLOBIOM_chart_hires.jpg
diff --git a/source/land_use/land.rst b/source/land_use/land.rst
index 5bec9c5..07c899d 100755
--- a/source/land_use/land.rst
+++ b/source/land_use/land.rst
@@ -2,7 +2,10 @@
 
 Land use change
 ---------------
-The model optimizes over six land cover types: cropland, grassland, short rotation plantations, managed forests, unmanaged forests and other natural land. Economic activities are associated with the first four land cover types. There are other three land cover types represented in the model: other agricultural land, wetlands, and not relevant (bare areas, water bodies, snow and ice, and artificial surfaces). These three categories are currently kept constant. Each Simulation Unit can contain the nine land cover types. The base year spatial distribution of land cover is based on the Global Land Cover 2000 (GLC2000). However, as any other global dataset of this type, GLC2000 suffers from large uncertainty (Fritz, See et al. 2011 :cite:`fritz_highlighting_2011`). Therefore auxiliary datasets and procedures are used to transform this “raw” data into a consistent dataset corresponding to the model needs.
+The model optimizes over six land cover types: cropland, grassland, short rotation plantations, managed forests, unmanaged forests and other natural land. Economic activities are associated with the first 
+four land cover types. There are other three land cover types represented in the model: other agricultural land, wetlands, and not relevant (bare areas, water bodies, snow and ice, and artificial surfaces). 
+These three categories are currently kept constant. Each Simulation Unit can contain the nine land cover types. The base year spatial distribution of land cover is based on the Global Land Cover 2000 (GLC2000). 
+However, as any other global dataset of this type, GLC2000 suffers from large uncertainty (Fritz et al., 2011 :cite:`fritz_highlighting_2011`). Therefore auxiliary datasets and procedures are used to transform this “raw” data into a consistent dataset corresponding to the model needs.
 
 .. _fig-globiom_land:
 
diff --git a/source/land_use/livestock.rst b/source/land_use/livestock.rst
index 5c8fabb..3a102c8 100755
--- a/source/land_use/livestock.rst
+++ b/source/land_use/livestock.rst
@@ -5,21 +5,37 @@ Livestock
 
 Livestock population
 ~~~~~~~~~~~~~~~~~~~~
-The principal variable characterizing the livestock production in GLOBIOM is the number of animals by species, production system and production type in each Simulation Unit. GLOBIOM differentiates four species aggregates: cattle and buffaloes (bovines), sheep and goats (small ruminants), pigs, and poultry. Eight production systems are specified for ruminants: grazing systems in arid (LGA), humid (LGH) and temperate/highland areas (LGT); mixed systems in arid (MXA), humid (MXH) and temperate/highland areas (MXT); urban systems (URB); and other systems (OTH). Mixed systems are an aggregate of the more detailed original Sere and Steinfeld’s classes (Sere and Steinfeld 1996 :cite:`sere_world_1996`) – mixed rainfed and mixed irrigated. Two production systems are specified for monogastrics: smallholders (SMH) and industrial systems (IND). In terms of production type, dairy and meat herds are modeled separately for ruminants: dairy herd includes adult females and replacement heifers, whose diets are distinguished. Poultry in smallholder systems is considered as mixed producer of meat and eggs, and poultry in industrial systems is split into laying hens and broilers, with differentiated diet regimes. Overall livestock numbers at the country level are, where possible while respecting minimum herd dynamics rules, harmonized with FAOSTAT. 
-
-The spatial distribution of ruminants and their allocation between production systems follows an updated version of Wint and Robinson (Wint and Robinson 2007 :cite:`wint_gridded_2007`). Since better information is not available, it is assumed that the share of dairy and meat herds within one region is the same in all production systems. The share is obtained from the FAO country level data about milk producing animals and total herd size. Monogastrics are not treated in a spatially explicit way since no reliable maps are currently available, and because monogastrics are not linked in the model to specific spatial features, like grasslands. The split between smallholder and industrial systems follows Herrero et al. (2013 :cite:`herrero_global_2013`).
+The principal variable characterizing the livestock production in GLOBIOM is the number of animals by species, production system and production type in each Simulation Unit. GLOBIOM differentiates 
+four species aggregates: cattle and buffaloes (bovines), sheep and goats (small ruminants), pigs, and poultry. Eight production systems are specified for ruminants: grazing systems in arid (LGA), 
+humid (LGH) and temperate/highland areas (LGT); mixed systems in arid (MXA), humid (MXH) and temperate/highland areas (MXT); urban systems (URB); and other systems (OTH). Mixed systems are an 
+aggregate of the more detailed original Sere and Steinfeld’s classes (Sere and Steinfeld, 1996 :cite:`sere_world_1996`) – mixed rainfed and mixed irrigated. Two production systems are specified 
+for monogastrics: smallholders (SMH) and industrial systems (IND). In terms of production type, dairy and meat herds are modeled separately for ruminants: dairy herd includes adult 
+females and replacement heifers, whose diets are distinguished. Poultry in smallholder systems is considered as mixed producer of meat and eggs, and poultry in industrial systems is split 
+into laying hens and broilers, with differentiated diet regimes. Overall livestock numbers at the country level are, where possible while respecting minimum herd dynamics rules, harmonized with FAOSTAT. 
+
+The spatial distribution of ruminants and their allocation between production systems follows an updated version of Wint and Robinson (Wint and Robinson, 2007 :cite:`wint_gridded_2007`). 
+Since better information is not available, it is assumed that the share of dairy and meat herds within one region is the same in all production systems. The share is obtained from the FAO 
+country level data about milk producing animals and total herd size. Monogastrics are not treated in a spatially explicit way since no reliable maps are currently available, and because 
+monogastrics are not linked in the model to specific spatial features, like grasslands. The split between smallholder and industrial systems follows Herrero et al. (2013) :cite:`herrero_global_2013`.
 
 Livestock products
 ~~~~~~~~~~~~~~~~~~
 Each livestock category is characterized by product yield, feed requirements, and a set of direct GHG emission coefficients. On the output side, seven products are defined: bovine meat and milk, small ruminant meat and milk, pig meat, poultry meat, and eggs. For each region, production type and production system, individual productivities are determined.
 
-Bovine and small ruminant productivities are estimated through the RUMINANT model (Herrero, Thornton et al. 2008, :cite:`herrero_systems_2008`; Herrero, Havlik et al. 2013 :cite:`herrero_global_2013`), in a three steps process which consists of first, specifying a plausible feed ration; second, calculating in RUMINANT the corresponding yield; and finally confronting at the region level with FAOSTAT (Supply Utilization Accounts) data on production. These three steps were repeated in a loop until a match with the statistical data was obtained. Monogastrics productivities were disaggregated from FAOSTAT based on assumptions about potential productivities and the relative differences in productivities between smallholder and industrial systems. The full detail of this procedure is provided in Herrero et al. (2013 :cite:`herrero_global_2013`).
+Bovine and small ruminant productivities are estimated through the RUMINANT model (Herrero et al., 2008 :cite:`herrero_systems_2008`; Herrero et al., 2013 :cite:`herrero_global_2013`), in a three steps process which 
+consists of first, specifying a plausible feed ration; second, calculating in RUMINANT the corresponding yield; and finally confronting at the region level with FAOSTAT (Supply Utilization Accounts) data on production. 
+These three steps were repeated in a loop until a match with the statistical data was obtained. Monogastrics productivities were disaggregated from FAOSTAT based on assumptions about potential productivities and the 
+relative differences in productivities between smallholder and industrial systems. The full detail of this procedure is provided in Herrero et al. (2013) :cite:`herrero_global_2013`.
 
 Final livestock products are expressed in primary commodity equivalents. Each product is considered as a differentiated good with a specific market except for bovine and small ruminant milk that are merged in a single milk market. The two milk types are therefore treated as perfect substitutes.
 
 Livestock feed
 ~~~~~~~~~~~~~~
-Feed requirements for ruminants are computed simultaneously with the yields (Herrero, Havlik et al. 2013 :cite:`herrero_global_2013`). Specific diets are defined for the adult dairy females, and for the other animals. The feed requirements are first calculated at the level of four aggregates – grains (concentrates), stover, grass, and other. When estimating the feed-yield couples, the RUMINANT model takes into account different qualities of these aggregates across regions and systems. Feed requirements for monogastrics are at this level determined through literature review presented in Herrero et al. (2013 :cite:`herrero_global_2013`). In general, it is assumed that in industrial systems pigs and poultry consume 10 and 12 kg dry matter of concentrates per TLU and day, respectively, and concentrates are the only feed sources. Smallholder animals get only one quarter of the amount of grains fed in industrial systems, the rest is supposed to come from other sources, like household waste, not explicitly represented in GLOBIOM.
+Feed requirements for ruminants are computed simultaneously with the yields (Herrero et al., 2013 :cite:`herrero_global_2013`). Specific diets are defined for the adult dairy females, and for the other animals. 
+The feed requirements are first calculated at the level of four aggregates – grains (concentrates), stover, grass, and other. When estimating the feed-yield couples, the RUMINANT model takes into account different 
+qualities of these aggregates across regions and systems. Feed requirements for monogastrics are at this level determined through literature review presented in Herrero et al. (2013) :cite:`herrero_global_2013`. 
+In general, it is assumed that in industrial systems pigs and poultry consume 10 and 12 kg dry matter of concentrates per TLU and day, respectively, and concentrates are the only feed sources. 
+Smallholder animals get only one quarter of the amount of grains fed in industrial systems, the rest is supposed to come from other sources, like household waste, not explicitly represented in GLOBIOM.
 
 The aggregate GRAINS input group is harmonized with feed quantities as reported at the country level in Commodity Balances of FAOSTAT. The harmonization proceeds in two steps, where first, GRAINS in the feed rations are adjusted so that total feed requirements at the country level match with total feed quantity in Commodity Balances, and second, “Grains” is disaggregated into 11 feed groups: Barley, Corn, Pulses, Rice, Sorghum & Millet, Soybeans, Wheat, Cereal Other, Oilseed Other, Crops Other, Animal Products. The adjustment of total GRAINS quantities is first done through shifts between the GRAINS and OTHER categories in ruminant systems. Hence, if total GRAINS are lower than the statistics, a part or total feed from the OTHER category is moved to GRAINS. If this is not enough, all GRAINS requirements of ruminants are shifted up in the same proportions. If total GRAINS are higher than the statistics, then firstly a part of them must be reallocated to the OTHER category. If this is not enough, values are to be kept, which then results in higher GRAINS demand than reported in FAOSTAT. This inconsistency is overcome in GLOBIOM, by creating a “reserve” of the missing GRAINS. This reserve is in simulations kept constant, thus it enables to reproduce the base year activity levels mostly consistent with FAOSTAT, but requires that all additional GRAINS demand arising over the simulation horizon is satisfied from real production. The decomposition of GRAINS into the 11 subcategories has to follow predefined minima and maxima of the shares of feedstuffs in a ration differentiated by species and region. At the same time, the shares of the feedstuffs corresponding to country level statistics need to be respected. This problem is solved as minimization of the square deviations from the prescribed minimum and maximum limits. In GLOBIOM, the balance between demand and supply of the crop products entering the GRAINS subcategories needs to be satisfied at regional level. Substitution ratios are defined for the byproducts of biofuel industry so that they can also enter the feed supply.
 
@@ -31,11 +47,24 @@ Finally, the feed aggregate OTHER is represented in a simplified way, where it i
 
 Grazing forage availability
 ~~~~~~~~~~~~~~~~~~~~~~~~~~~
-The demand and supply of grass need to match at the level of Simulation Unit in GLOBIOM. But reliable information about grass forage supply is not available even at the country level. The forage supply is a product of the utilized grassland area and of forage productivity. However, at global scale, Ramankutty et al. (2008 :cite:`ramankutty_farming_2008`) estimated that the extent of pastures spans in the 90% confidence interval between 2.36 and 3.00 billion hectares. The FAOSTAT estimate of 3.44 billion hectares itself falls outside of this interval which illustrates the level of uncertainty in the grassland extent. Similarly, with respect to forage productivity, different grassland production models perform better for different forage production systems and all are confronted with considerable uncertainty due to limited information about vegetation types, management practices, etc. (Conant and Paustian 2004 :cite:`conant_grassland_2004`). These limitations precluded reliance on any single source of information or output from a single model. Therefore three different grass productivity sources were considered: CENTURY on native grasslands, CENTURY on native and managed grasslands, and EPIC on managed grasslands.  
-
-A systematic process was developed for selecting the suitable productivity source for each of GLOBIOM’s 30 regions. This process allowed reliance on sound productivity estimates that are consistent with other GLOBIOM datasets like spatial livestock distribution and feed requirements. Within this selection process, the area of utilized grasslands corresponding to the base year 2000 was determined simultaneously with the suitable forage productivity layer. Two selection criteria were used: livestock requirements for forage and area of permanent meadows and pastures from FAOSTAT. The selection process was based on simultaneous minimization of  i) the difference between livestock demand for forage and the model-estimates of forage supply and ii) the difference between the utilized grassland area and FAOSTAT statistics on permanent meadows and pastures. Regional differentiation in grassland management intensity, ranging from dry grasslands with minimal inputs to mesic, planted pastures that are intensively managed with large external inputs – further informed the model selection by enabling constraints in the number of models for dry grasslands.
-
-To calculate the utilized grassland area, the potential grassland area was first defined as the area belonging to one of the following GLC2000 land cover classes: 13 (Herbaceous Cover, closed-open), 16-18 (Cultivated and managed areas, Mosaic: Cropland / Tree Cover / Other natural vegetation, Mosaic: Cropland / Shrub and/or grass cover), excluding area identified as cropland according to the IFPRI crop distribution map (You and Wood 2006 :cite:`you_entropy_2006`), and 11, 12, 14 (Shrub Cover, closed-open, evergreen, Shrub Cover, closed-open, deciduous, Sparse herbaceous or sparse shrub cover). In each Simulation Unit the utilized area was calculated by dividing total forage requirements by forage productivity. In Simulation Units where utilized area was smaller than the potential grassland area, the difference would be allocated to either “Other Natural Land” or “Other Agricultural Land” depending on the underlying GLC2000 class. In Simulation Units where the grassland area necessary to produce the forage required in the base year was larger than the potential grassland area, a “reserve” was created to ensure base year feasibility, but all the additional grass demand arising through future livestock production increases needed to be satisfied from grasslands.
+The demand and supply of grass need to match at the level of Simulation Unit in GLOBIOM. But reliable information about grass forage supply is not available even at the country level. 
+The forage supply is a product of the utilized grassland area and of forage productivity. However, at global scale, Ramankutty et al. (2008) :cite:`ramankutty_farming_2008` estimated that 
+the extent of pastures spans in the 90% confidence interval between 2.36 and 3.00 billion hectares. The FAOSTAT estimate of 3.44 billion hectares itself falls outside of this interval 
+which illustrates the level of uncertainty in the grassland extent. Similarly, with respect to forage productivity, different grassland production models perform better for different 
+forage production systems and all are confronted with considerable uncertainty due to limited information about vegetation types, management practices, etc. (Conant and Paustian, 2004 :cite:`conant_grassland_2004`). 
+These limitations precluded reliance on any single source of information or output from a single model. Therefore three different grass productivity sources were considered: CENTURY on native grasslands, 
+CENTURY on native and managed grasslands, and EPIC on managed grasslands.  
+
+A systematic process was developed for selecting the suitable productivity source for each of GLOBIOM’s 30 regions. This process allowed reliance on sound productivity estimates that are consistent with 
+other GLOBIOM datasets like spatial livestock distribution and feed requirements. Within this selection process, the area of utilized grasslands corresponding to the base year 2000 was determined simultaneously 
+with the suitable forage productivity layer. Two selection criteria were used: livestock requirements for forage and area of permanent meadows and pastures from FAOSTAT. The selection process was based on 
+simultaneous minimization of  i) the difference between livestock demand for forage and the model-estimates of forage supply and ii) the difference between the utilized grassland area and FAOSTAT statistics on 
+permanent meadows and pastures. Regional differentiation in grassland management intensity, ranging from dry grasslands with minimal inputs to mesic, planted pastures that are intensively managed with large external 
+inputs – further informed the model selection by enabling constraints in the number of models for dry grasslands.
+
+To calculate the utilized grassland area, the potential grassland area was first defined as the area belonging to one of the following GLC2000 land cover classes: 13 (Herbaceous Cover, closed-open), 16-18 
+(Cultivated and managed areas, Mosaic: Cropland / Tree Cover / Other natural vegetation, Mosaic: Cropland / Shrub and/or grass cover), excluding area identified as cropland according to the IFPRI crop distribution map 
+(You and Wood, 2006 :cite:`you_entropy_2006`), and 11, 12, 14 (Shrub Cover, closed-open, evergreen, Shrub Cover, closed-open, deciduous, Sparse herbaceous or sparse shrub cover). In each Simulation Unit the utilized area was calculated by dividing total forage requirements by forage productivity. In Simulation Units where utilized area was smaller than the potential grassland area, the difference would be allocated to either “Other Natural Land” or “Other Agricultural Land” depending on the underlying GLC2000 class. In Simulation Units where the grassland area necessary to produce the forage required in the base year was larger than the potential grassland area, a “reserve” was created to ensure base year feasibility, but all the additional grass demand arising through future livestock production increases needed to be satisfied from grasslands.
 
 .. _fig-forage:
 .. figure:: /_static/GLOBIOM_forage_availability.png
@@ -43,9 +72,20 @@ To calculate the utilized grassland area, the potential grassland area was first
 
    Data sources used to parameterize forage availability in different world regions. CENTURY_NAT – CENTURY model for native grasslands; CENTURY_MGT – CENTURY model for productive grasslands; EPIC_EXT – EPIC model for grasslands under extensive management; EPIC_MID – EPIC model for grasslands under semi-intensive management; EPIC_INT – EPIC model for grasslands under intensive management.
    
-Forage productivity was estimated using the CENTURY (Parton, Schimel et al. 1987 :cite:`parton_analysis_1987`; Parton, Scurlock et al. 1993 :cite:`parton_observations_1993`) and EPIC (Williams and Singh 1995 :cite:`williams_computer_1995`) models. The CENTURY model was run globally at 0.5 degree resolution to estimate native forage and browse and planted pastures productivity. It was initiated with 2000 year spin-ups using mean monthly climate from the Climate Research Unit (CRU) of the University of East Anglia with native vegetation for each grid cell, except cells dominated by rock, ice, and water, which were excluded. Information about native vegetation was derived from the Potsdam intermodal comparison study (Schloss, Kicklighter et al. 1999 :cite:`schloss_comparing_1999`). Plant community and land management (grazing) was based on growing-season grazing and 50 per cent forage removal. Areas under native vegetation that were grazed were identified using the map of native biomes subject to grazing and subtracting estimated crop area within those biomes in 2006 (Ramankutty, Evan et al. 2008 :cite:`ramankutty_farming_2008`). It is assumed 50 per cent grazing efficiency for grass, and 25 per cent for browse for native grasslands. These CENTURY-based estimates of native grassland forage production (CENTURY_NAT) were used for most regions with low-productivity grasslands (:numref:`fig-forage`). 
-
-Both the CENTURY and EPIC models were used to estimate forage production in mesic, more productive regions. For the CENTURY model, forage yield was simulated using a highly-productive, warm-season grass parameterization. Production was modeled in all cells and applied to areas of planted pasture, which were estimated based on biomes that were not native rangelands, but were under pasture in 2006 according to Ramankutty (Ramankutty, Evan et al. 2008 :cite:`ramankutty_farming_2008`). Pastures were replanted in the late winter every ten years, with grazing starting in the second year. Observed monthly precipitation and minimum and maximum temperatures between 1901 and 2006 were from the CRU Time Series data, CRU TS30 (Mitchell and Jones 2005 :cite:`mitchell_improved_2005`) Soils data were derived from the FAO Soil Map of the World, as modified by Reynolds, Jackson et al. (2000 :cite:`reynolds_estimating_2000`). CENTURY model output for productive pastures (CENTURY_MGT) were the best-match for area/forage demand in much of the world with a mixture of mesic and drier pastures.
+Forage productivity was estimated using the CENTURY (Parton et al., 1987 :cite:`parton_analysis_1987`; Parton et al., 1993 :cite:`parton_observations_1993`) and EPIC (Williams and Singh, 1995 :cite:`williams_computer_1995`) 
+models. The CENTURY model was run globally at 0.5 degree resolution to estimate native forage and browse and planted pastures productivity. It was initiated with 2000 year spin-ups using mean monthly climate from the 
+Climate Research Unit (CRU) of the University of East Anglia with native vegetation for each grid cell, except cells dominated by rock, ice, and water, which were excluded. Information about native vegetation was derived 
+from the Potsdam intermodal comparison study (Schloss et al., 1999 :cite:`schloss_comparing_1999`). Plant community and land management (grazing) was based on growing-season grazing and 50 per cent forage removal. 
+Areas under native vegetation that were grazed were identified using the map of native biomes subject to grazing and subtracting estimated crop area within those biomes in 2006 
+(Ramankutty et al., 2008 :cite:`ramankutty_farming_2008`). It is assumed 50 per cent grazing efficiency for grass, and 25 per cent for browse for native grasslands. These CENTURY-based estimates of native grassland 
+forage production (CENTURY_NAT) were used for most regions with low-productivity grasslands (:numref:`fig-forage`). 
+
+Both the CENTURY and EPIC models were used to estimate forage production in mesic, more productive regions. For the CENTURY model, forage yield was simulated using a highly-productive, warm-season grass 
+parameterization. Production was modeled in all cells and applied to areas of planted pasture, which were estimated based on biomes that were not native rangelands, but were under pasture in 
+2006 according to Ramankutty (Ramankutty et al., 2008 :cite:`ramankutty_farming_2008`). Pastures were replanted in the late winter every ten years, with grazing starting in the second year. 
+Observed monthly precipitation and minimum and maximum temperatures between 1901 and 2006 were from the CRU Time Series data, CRU TS30 (Mitchell and Jones, 2005 :cite:`mitchell_improved_2005`) 
+Soils data were derived from the FAO Soil Map of the World, as modified by Reynolds et al. (2000) :cite:`reynolds_estimating_2000`. CENTURY model output for productive pastures (CENTURY_MGT) were 
+the best-match for area/forage demand in much of the world with a mixture of mesic and drier pastures.
 
 .. _fig-forage-livestock:
 .. figure:: /_static/GLOBIOM_forage_livestock.png
@@ -53,13 +93,23 @@ Both the CENTURY and EPIC models were used to estimate forage production in mesi
 
    Forage available for livestock in tonnes of dry matter per hectare as the result of combination of outputs from the CENTURY and EPIC models.
  
-The EPIC model was the best fit for much of Europe and Eastern Asia, where most of the forage production is in intensively-managed grasslands. The EPIC simulations used the same soil and climatic drivers as the CENTURY runs plus topography data (high-resolution global Shuttle Radar Topography Mission digital elevation model (SRTM) and the Global 30 Arc Second Elevation Data (GTOPO30). Warm and cold seasonal grasses were simulated in EPIC, and the simulations included a range of management intensities represented by different levels of nitrogen fertilizer inputs and off-take rates. The most intensive management minimizing nitrogen stress and applying 80% off-take rates (EPIC_INT) was found to be the best match for South Korea. Highly fertilized grasslands but with an off-take rate of 50% only were identified in Western Europe, China and Japan (EPIC_MID), and finally extensive management, only partially satisfying the nitrogen requirements and considering 20% off-take rates corresponded best to Central and Northern Europe and South-East Asia (EPIC_EXT). The resulting hybrid forage availability map is represented in :numref:`fig-forage-livestock`.
+The EPIC model was the best fit for much of Europe and Eastern Asia, where most of the forage production is in intensively-managed grasslands. The EPIC simulations used the same soil and climatic drivers as the 
+CENTURY runs plus topography data (high-resolution global Shuttle Radar Topography Mission digital elevation model (SRTM) and the Global 30 Arc Second Elevation Data (GTOPO30). 
+Warm and cold seasonal grasses were simulated in EPIC, and the simulations included a range of management intensities represented by different levels of nitrogen fertilizer inputs and off-take rates. 
+The most intensive management minimizing nitrogen stress and applying 80% off-take rates (EPIC_INT) was found to be the best match for South Korea. Highly fertilized grasslands but with an off-take rate 
+of 50% only were identified in Western Europe, China and Japan (EPIC_MID), and finally extensive management, only partially satisfying the nitrogen requirements and considering 20% off-take rates corresponded best 
+to Central and Northern Europe and South-East Asia (EPIC_EXT). The resulting hybrid forage availability map is represented in :numref:`fig-forage-livestock`.
 
 Livestock dynamics
 ~~~~~~~~~~~~~~~~~~
-In general, the number of animals of a given species and production type in a particular production system and Supply Unit is an endogenous variable. This means that it will decrease or increase in relation to changes in demand and the relative profitability with respect to competing activities.
+In general, the number of animals of a given species and production type in a particular production system and Supply Unit is an endogenous variable. This means that it will decrease or increase in relation to changes in 
+demand and the relative profitability with respect to competing activities.
 
-Herd dynamics constraints need however to be respected. First, dairy herds are constituted of adult females and followers, and expansion therefore occurs in predefined proportions in the two groups. Moreover, for regions where the specialized meat herds are insignificant (no suckler cows), expansion of meat animals (surplus heifers and males) is also assumed proportional in size to the dairy herd. The ruminants in urban systems are not allowed to expand because this category is not well known and because it is fairly constrained by available space in growing cities. Finally, the decrease of animals per system and production type higher than 15 per cent per 10 years period are not considered, and no increase by more than 100 per cent on the same period. At the level of individual systems, the decrease can however be as deep as 50 per cent per system on a single period.
+Herd dynamics constraints need however to be respected. First, dairy herds are constituted of adult females and followers, and expansion therefore occurs in predefined proportions in the two groups. Moreover, 
+for regions where the specialized meat herds are insignificant (no suckler cows), expansion of meat animals (surplus heifers and males) is also assumed proportional in size to the dairy herd. 
+The ruminants in urban systems are not allowed to expand because this category is not well known and because it is fairly constrained by available space in growing cities. Finally, the decrease 
+of animals per system and production type higher than 15 per cent per 10 years period are not considered, and no increase by more than 100 per cent on the same period. At the level of individual systems, 
+the decrease can however be as deep as 50 per cent per system on a single period.
 
-For monogastrics, the assumption is made that all additional supply will come from industrial systems and hence the number of animals in other systems is kept constant (Keyzer, Merbis et al. 2005 :cite:`keyzer_diet_2005`). 
+For monogastrics, the assumption is made that all additional supply will come from industrial systems and hence the number of animals in other systems is kept constant (Keyzer et al., 2005 :cite:`keyzer_diet_2005`). 
 
diff --git a/source/land_use/spatial.rst b/source/land_use/spatial.rst
index e36d6bf..fac04a6 100755
--- a/source/land_use/spatial.rst
+++ b/source/land_use/spatial.rst
@@ -2,6 +2,18 @@
 
 Spatial resolution
 -----------------------
-Land resources and their characteristics are the fundamental elements of the GLOBIOM modelling approach. In order to enable global bio-physical process modelling of agricultural and forest production, a comprehensive database has been built (Skalsky, Tarasovicova et al. 2008 :cite:`skalsky_geo-bene_2008`), which contains geo-spatial data on soil, climate/weather, topography, land cover/use, and crop management (e.g. fertilization, irrigation). The data were compiled from various sources (FAO, ISRIC, USGS, NASA, CRU UEA, JRC, IFRPI, IFA, WISE, etc.) and significantly vary with respect to spatial, temporal, and attribute resolutions, thematic relevance, accuracy, and reliability. Therefore, data were harmonized into several common spatial resolution layers including 5 and 30 Arcmin as well as country layers. Subsequently, Homogeneous Response Units (HRU) have been delineated by geographically clustering according to only those parameters of the landscape, which are generally not changing over time and are thus invariant with respect to land use and management or climate change. At the global scale, five altitude classes, seven slope classes, and five soil classes have been included. In a second step, the HRU layer is intersected with a 0.5 x 0.5 degree grid and country boundaries to delineate Simulation Units (SimUs) which contain other relevant information such as global climate data, land category/use data, irrigation data, etc. In total, 212,707 SimUs are delineated by clustering 5 x 5 minutes of arc pixels according to five criteria: altitude, slope, and soil class, 0.5 x 0.5 degrees grid, and the country boundaries. The SimUs are the basis for estimation of land use/management parameters in all other supporting models as well. For each SimU a number of land management options are simulated using the bio-physical process model EPIC (Environmental Policy Integrated Climate Model; Izaurralde et al., 2006 :cite:`izaurralde_simulating_2006`; Williams and Singh, 1995; :cite:`williams_computer_1995`). For the SSP application of GLOBIOM, in order to ease computation time, the input data sets and the model resolution were agregated to 2 x 2 degree cells disaggregated only by country boundaries and by three agro-ecological zones used in the livestock production system classification: arid, humid, temperate and tropical highlands. This led to a total of 10,894 different Supply Units.
+Land resources and their characteristics are the fundamental elements of the GLOBIOM modelling approach. In order to enable global bio-physical process modelling of agricultural and forest production, a 
+comprehensive database has been built (Skalsky et al., 2008 :cite:`skalsky_geo-bene_2008`), which contains geo-spatial data on soil, climate/weather, topography, land cover/use, and crop management 
+(e.g. fertilization, irrigation). The data were compiled from various sources (FAO, ISRIC, USGS, NASA, CRU UEA, JRC, IFRPI, IFA, WISE, etc.) and significantly vary with respect to spatial, temporal, 
+and attribute resolutions, thematic relevance, accuracy, and reliability. Therefore, data were harmonized into several common spatial resolution layers including 5 and 30 Arcmin as well as country layers. 
+Subsequently, Homogeneous Response Units (HRU) have been delineated by geographically clustering according to only those parameters of the landscape, which are generally not changing over time and are thus 
+invariant with respect to land use and management or climate change. At the global scale, five altitude classes, seven slope classes, and five soil classes have been included. 
+
+In a second step, the HRU layer is intersected with a 0.5 x 0.5 degree grid and country boundaries to delineate Simulation Units (SimUs) which contain other relevant information such as global climate data, 
+land category/use data, irrigation data, etc. In total, 212,707 SimUs are delineated by clustering 5 x 5 minutes of arc pixels according to five criteria: altitude, slope, and soil class, 0.5 x 0.5 
+degrees grid, and the country boundaries. The SimUs are the basis for estimation of land use/management parameters in all other supporting models as well. For each SimU a number of land management 
+options are simulated using the bio-physical process model EPIC (Environmental Policy Integrated Climate) (Izaurralde et al., 2006 :cite:`izaurralde_simulating_2006`; Williams and Singh, 1995 :cite:`williams_computer_1995`). 
+For the SSP application of GLOBIOM, in order to ease computation time, the input data sets and the model resolution were agregated to 2 x 2 degree cells disaggregated only by country boundaries and by three agro-ecological 
+zones used in the livestock production system classification: arid, humid, temperate and tropical highlands. This led to a total of 10,894 different Supply Units.
 
  
diff --git a/source/overview/index.rst b/source/overview/index.rst
index d27eb09..561eb02 100755
--- a/source/overview/index.rst
+++ b/source/overview/index.rst
@@ -4,7 +4,7 @@ Overview
 ==============
 The IIASA IAM framework consists of a combination of five different models or modules - the energy model MESSAGE, the land use model GLOBIOM, the air pollution and GHG  model GAINS, the aggregated macro-economic model MACRO and the simple climate model MAGICC - which complement each other and are specialized in different areas. All models and modules together build the IIASA IAM framework, also referred to as MESSAGE-GLOBIOM owing to the fact that the energy model MESSAGE and the land use model  GLOBIOM are its most important components. The five models provide input to and iterate between each other during a typical scenario development cycle. Below is a brief overview of how the models interact with each other, specifically in the context of developing the SSP scenarios.
 
-MESSAGE represents the core of the IIASA IAM framework (:numref:`fig-iiasaiam`) and its main task is to optimize the energy system so that it can satisfy specified energy demands at the lowest costs. MESSAGE carries out this optimization in an iterative setup with MACRO, which provides estimates of the macro-economic demand response that results from energy system and services costs computed by MESSAGE. For the six commercial end-use demand categories depicted in MESSAGE (see :ref:`demand`), MACRO will adjust useful energy demands, until the two models have reached equilibrium (see :ref:`macro`). This iteration reflects price-induced energy efficiency improvements that can occur when energy prices change.
+MESSAGE (Huppmann et al., 2019 :cite:`huppmann_message_2019`) represents the core of the IIASA IAM framework (:numref:`fig-iiasaiam`) and its main task is to optimize the energy system so that it can satisfy specified energy demands at the lowest costs. MESSAGE carries out this optimization in an iterative setup with MACRO, which provides estimates of the macro-economic demand response that results from energy system and services costs computed by MESSAGE. For the six commercial end-use demand categories depicted in MESSAGE (see :ref:`demand`), MACRO will adjust useful energy demands, until the two models have reached equilibrium (see :ref:`macro`). This iteration reflects price-induced energy efficiency improvements that can occur when energy prices change.
 
 The scientific software powering the global MESSAGE model is called the |MESSAGEix| framework, an open-source, versatile implementation of a linear optimization problem, with the option of coupling to a computable general equilibrium (CGE) model to incorporate the effect of price changes on economic activity and demand for commodities and resources. |MESSAGEix| is integrated with the *ix modeling platform* (ixmp), a powerful "data warehouse" for version control of reference timeseries, input data and model results. ixmp provides interfaces to the scientific programming languages Python and R for efficient, scripted workflows for data processing and visualisation of results (Huppmann et al., 2019 :cite:`huppmann_2019_MESSAGEix`).
 
@@ -12,14 +12,14 @@ GLOBIOM provides MESSAGE with information on land use and its implications, incl
 
 Air pollution implications of the energy system are accounted for in MESSAGE by applying technology-specific air pollution coefficients from GAINS (see :ref:`gains`).
 
-In general, cumulative global carbon emissions from all sectors are constrained at different levels, with equivalent pricing applied to other GHGs, to reach the desired radiative forcing levels (cf. right-hand side :numref:`fig-iiasaiam`). The climate constraints are thus taken up in the coupled MESSAGE-GLOBIOM optimization, and the resulting carbon price is fed back to the full-fledged GLOBIOM model for full consistency. Finally, the combined results for land use, energy, and industrial emissions from MESSAGE and GLOBIOM are merged and fed into MAGICC (see :ref:`magicc`), a global carbon-cycle and climate model, which then provides estimates of the climate implications in terms of atmospheric concentrations, radiative forcing, and global-mean temperature increase. Importantly, climate impacts and impacts of the carbon cycle are currently not accounted for in the IIASA IAM framework. The entire framework is linked to an online database infrastructure which allows straightforward visualisation, analysis, comparison and dissemination of results (`Riahi et al., 2016 <http://pure.iiasa.ac.at/13280/>`_ :cite:`riahi_shared_2016`).
+In general, cumulative global carbon emissions from all sectors are constrained at different levels, with equivalent pricing applied to other GHGs, to reach the desired radiative forcing levels (cf. right-hand side :numref:`fig-iiasaiam`). The climate constraints are thus taken up in the coupled MESSAGE-GLOBIOM optimization, and the resulting carbon price is fed back to the full-fledged GLOBIOM model for full consistency. Finally, the combined results for land use, energy, and industrial emissions from MESSAGE and GLOBIOM are merged and fed into MAGICC (see :ref:`magicc`), a global carbon-cycle and climate model, which then provides estimates of the climate implications in terms of atmospheric concentrations, radiative forcing, and global-mean temperature increase. Importantly, climate impacts and impacts of the carbon cycle are currently not accounted for in the IIASA IAM framework. The entire framework is linked to an online database infrastructure which allows straightforward visualisation, analysis, comparison and dissemination of results (`Riahi et al., 2017 <http://pure.iiasa.ac.at/13280/>`_ :cite:`riahi_shared_2017`).
 
 .. _fig-iiasaiam:
 
 .. figure:: /_static/iiasaiam.png
    :width: 900px
 
-   Overview of the IIASA IAM framework. Coloured boxes represent respective specialized disciplinary models which are integrated for generating internally consistent scenarios (Fricko et al., 2016 :cite:`fricko_marker_2016`). 
+   Overview of the IIASA IAM framework. Coloured boxes represent respective specialized disciplinary models which are integrated for generating internally consistent scenarios (Fricko et al., 2017 :cite:`fricko_marker_2017`). 
 
 .. toctree::
    :maxdepth: 1
diff --git a/source/overview/spatial.rst b/source/overview/spatial.rst
index a7f8534..7ee3776 100755
--- a/source/overview/spatial.rst
+++ b/source/overview/spatial.rst
@@ -10,7 +10,7 @@ The combined MESSAGE-GLOBIOM framework has global coverage and divides the world
 
    Map of 11 MESSAGE-GLOBIOM regions inclduing their aggregation to the four regions used in the Representative Concentration Pathways (RCPs).
 
-The country definitions of the 11 MESSAGE regions are described in the table below (:numref:`tab-reg`).
+The country definitions of the 11 MESSAGE regions are described in the table below (:numref:`tab-reg`). In some scenarios, the MESSAGE region of FSU is disaggregated into four sub-regions resulting in a 14-region MESSAGE model.
 
 .. _tab-reg:
 .. list-table:: Listing of 11 regions used in MESSAGE-GLOBIOM, including their country definitions.
diff --git a/source/overview/temporal.rst b/source/overview/temporal.rst
index 6763e5f..d0c70e5 100755
--- a/source/overview/temporal.rst
+++ b/source/overview/temporal.rst
@@ -1,7 +1,7 @@
 Time steps
 =================
-MESSAGE models the time horizon 1990 to 2110 in 5- and 10 year time steps where the first 5 periods (1990, 1995, 2000, 2005, 2010) are 5-year periods and the remaining 10 periods are 10-year periods (2020, 2030, 2040, 2050, 2060, 2070, 2080, 2090, 2100, 2110). The first four periods up to 2005 are fully calibrated, i.e. the model has no flexibility to change in these five periods. The 2010 period is partly calibrated so far, some recent trends are included in this time period, but some flexibility remains. In scenario applications the 2010 period is typically fixed to its baseline development so that future climate and energy policy cannot induce changes in the past. The reporting years are the final years of periods which implies that investments that lead to the capacities in the reporting year are the average annual investments over the entire period the reporting year belongs to.
+MESSAGE models the time horizon 2010 to 2110 generally in 10-year periods (2020, 2030, 2040, 2050, 2060, 2070, 2080, 2090, 2100, 2110), using 2010 as the base year. The 2020 period is partly calibrated so far, some recent trends are included in this time period, but some flexibility remains. The reporting years are the final years of periods which implies that investments that lead to the capacities in the reporting year are the average annual investments over the entire period the reporting year belongs to. In some model versions, the model has been calibrated to 2015 running with 5-year modeling periods by the middle of the Century (2020, 2025, 2030, 2035, 2040, 2045, 2050, 2055, 2060) and 10-year periods between 2060 and 2110.
 
-MESSAGE can both operate perfect foresight over the entire time horizon, limited foresight (e.g., two or three periods into the future) or myopically, optimizing one period at a time (Keppo and Strubegger, 2010 :cite:`keppo_short_2010`). Most frequently MESSAGE is run with perfect foresight, but for specific applications such as delayed participation in a global climate regime without anticipation (Krey and Riahi, 2009 :cite:`krey_implications_2009`; O'Neill et al., 2010 :cite:`oneill_mitigation_2010`) limited foresight is used.
+MESSAGE can both operate perfect foresight over the entire time horizon, limited foresight (e.g., two or three periods into the future) or myopically, optimizing one period at a time (Keppo and Strubegger, 2010 :cite:`keppo_short_2010`) (see `Mathematical Specification <https://message.iiasa.ac.at/en/stable/model/MESSAGE/model_solve.html#recursive-dynamic-and-myopic-model>`_ for more details). Most frequently MESSAGE is run with perfect foresight, but for specific applications such as delayed participation in a global climate regime without anticipation (Krey and Riahi, 2009 :cite:`krey_implications_2009`; O'Neill et al., 2010 :cite:`oneill_mitigation_2010`) limited foresight is used.
 
 GLOBIOM models the time horizon 2000 to 2100 in 10 year time steps (2000, 2010, 2020, 2030, 2040, 2050, 2060, 2070, 2080, 2090, 2100) with the year 2000 being the base year of the model.  The model is recursive-dynamic, i.e. it is solved for each period individually and then passes on results to the subsequent periods. The linkage between MESSAGE and GLOBIOM relies on the model results of the periods 2020 to 2100.
\ No newline at end of file
diff --git a/source/socio_econ/narratives.rst b/source/socio_econ/narratives.rst
index 27998ae..1e5362d 100755
--- a/source/socio_econ/narratives.rst
+++ b/source/socio_econ/narratives.rst
@@ -2,7 +2,7 @@
 
 SSP narratives
 ===============
-Narratives have been developed for the Shared Socioeconomic Pathways (SSPs) (O’Neill et al., 2015 :cite:`oneill_roads_2015`). These descriptions of alternative futures of societal development span a range of possible worlds that stretch along two climate-change-related dimensions: mitigation and adaptation challenges. The SSPs reflect five different developments of the world that are characterized by varying levels of global challenges (see `Riahi et al., 2016 <http://pure.iiasa.ac.at/13280/>`_  :cite:`riahi_shared_2016` for an overview). In the following, the three narratives that have been translated into quantitative scenarios with MESSAGEix-GLOBIOM are presented (Fricko et al., 2016 :cite:`fricko_marker_2016`):
+Narratives have been developed for the Shared Socioeconomic Pathways (SSPs) (O’Neill et al., 2015 :cite:`oneill_roads_2015`). These descriptions of alternative futures of societal development span a range of possible worlds that stretch along two climate-change-related dimensions: mitigation and adaptation challenges. The SSPs reflect five different developments of the world that are characterized by varying levels of global challenges (see `Riahi et al., 2017 <http://pure.iiasa.ac.at/13280/>`_  :cite:`riahi_shared_2017` for an overview). In the following, the three narratives that have been translated into quantitative scenarios with MESSAGE-GLOBIOM are presented (Fricko et al., 2017 :cite:`fricko_marker_2017`):
 
 SSP1 Narrative: Sustainability — Taking the green road
 ----------------------------------------------------
diff --git a/source/socio_econ/pop_GDP.rst b/source/socio_econ/pop_GDP.rst
index 0e3788f..4f9778d 100755
--- a/source/socio_econ/pop_GDP.rst
+++ b/source/socio_econ/pop_GDP.rst
@@ -2,6 +2,14 @@ Population and GDP
 ====================
 Population and economic developments have strong implications for the anticipated mitigation and adaptation challenges. For example, a larger, poorer and less educated population will have more difficulties to adapt to the detrimental effects of climate change (O’Neill et al., 2014 :cite:`oneill_new_2014`). The primary drivers of future energy demand in MESSAGEix are projections of total population and GDP at purchasing power parity exchange rates, denoted as GDP (PPP). In addition to total population, the urban/rural split of population is relevant for the `MESSAGEix-Access <http://www.iiasa.ac.at/web/home/research/researchPrograms/Energy/MESSAGE-Access.en.html>`_ version of the model which distinguishes rural and urban population with different household incomes in developing country regions.
 
-Understanding how population and economic growth develops in the SSPs gives a first layer of understanding of the multiple mitigation and adaptation challenges. Population growth evolves in response to how fertility, mortality, migration, and education of various social strata are assumed to change over time. In SSP2, global population peaks at 9.4 billion people around 2070, and slowly declines thereafter (KC and Lutz, 2015 :cite:`kc_human_2014`). Gross Domestic Product (GDP) follows regional historical trends (Dellink et al., 2015 :cite:`dellink_long-term_2015`). In SSP2, average income grows by a factor of six and reaches about 60,000 USD/capita by the end of the century (all GDP/capita figures use USD2005 and purchasing-power-parity – PPP). The SSP2 GDP projection is situated in-between the estimates for SSP1 and SSP3, which reach global average income levels of 82,000 USD2005 and 22,000 USD2005, respectively, by the end of the century. SSP2 depicts a future of global progress where developing countries achieve significant economic growth. Today, average per capita income in the global North is about five times higher than in the global South. In SSP2, developing countries reach today’s average income levels of the OECD between 2060 and 2090, depending on the region. However, modest improvements of educational attainment levels result in declines in education-specific fertility rates, leading to incomplete economic convergence across different world regions. This is particularly an issue for Africa. Overall, both the population and GDP developments in SSP2 (Fricko et al., 2016 :cite:`fricko_marker_2016`) are designed to be situated in the middle of the road between SSP1 and SSP3, see KC and Lutz (2015) :cite:`kc_human_2014` and Dellink et al (2015) :cite:`dellink_long-term_2015` for details.
+Understanding how population and economic growth develops in the SSPs gives a first layer of understanding of the multiple mitigation and adaptation challenges. Population growth evolves in response to how fertility, 
+mortality, migration, and education of various social strata are assumed to change over time. In SSP2, global population peaks at 9.4 billion people around 2070, and slowly declines thereafter 
+(KC and Lutz, 2015 :cite:`kc_human_2014`). Gross Domestic Product (GDP) follows regional historical trends (Dellink et al., 2015 :cite:`dellink_long-term_2015`). In SSP2, average income grows by a factor of six and reaches 
+about 60,000 USD/capita by the end of the century (all GDP/capita figures use USD2005 and purchasing-power-parity – PPP). The SSP2 GDP projection is situated in-between the estimates for SSP1 and SSP3, 
+which reach global average income levels of 82,000 USD2005 and 22,000 USD2005, respectively, by the end of the century. SSP2 depicts a future of global progress where developing countries achieve significant economic growth. 
+Today, average per capita income in the global North is about five times higher than in the global South. In SSP2, developing countries reach today’s average income levels of the OECD between 2060 and 2090, depending on 
+the region. However, modest improvements of educational attainment levels result in declines in education-specific fertility rates, leading to incomplete economic convergence across different world regions. 
+This is particularly an issue for Africa. Overall, both the population and GDP developments in SSP2 are designed to be situated in the middle of the road between SSP1 and SSP3, see 
+KC and Lutz (2015) :cite:`kc_human_2014`, Dellink et al. (2015) :cite:`dellink_long-term_2015` and Fricko et al. (2017) :cite:`fricko_marker_2017` for more details.
 
 The full quantitative data set of demographic and economic projections for the SSPs can be found in an online database (`SSP database <https://tntcat.iiasa.ac.at/SspDb/dsd?Action=htmlpage&page=welcome>`_).