On April 29, according to data graph, the bottom anemometer (wind2) became misaligned (by ~ 90 deg) relative to top (wind1) anemometer. 