Covariability Found in Evapotranspiration and Satellite-derived Vegetation Index Interannual Change Over Northern Asia

The transpiration from the vegetation of extensive forests plays an important role in the water and energy cycle in the climate system. We reported that the continental scale evapotranspiration (ET) shows similar interannual change with that of the satellite-derived Normalized Difference Vegetation Index (NDVI) from 1982 to 2000 (Suzuki et al., AGU 2002 fall meeting). This study develops it by examining ET interannual change over northern Asia, and comparing with two essential climate parameters, temperature and precipitation, in addition to the NDVI . Monthly global NDVI data were acquired from the Pathfinder AVHRR Land data (1 x 1 degree). The ET was estimated by so-called ``atmospheric water budget method, '' that is, a method for estimating the ET from the precipitation and water vapor divergence in the air column on each grid cell. This study used 2.5 x 2.5 degree spatial resolution atmospheric data created by the National Centers for Environmental Prediction, and gridded (2.5 x 2.5 degree) precipitation data provided by the CPC Merged Analysis of Precipitation. Temperature data are obtained from CRU TS2.0 dataset which consists of monthly gridded (0.5 x 0.5 degree) value based on the surface station observation. The monthly anomaly of the NDVI, ET, temperature, and precipitation was calculated for each 2.5 x 2.5 degree grid cell from 1982 to 2000, and year-to-year variation of the ET was compared with other three parameters. In addition to the significant positive correlation between the ET and NDVI interannual changes, high positive ET-temperature and ET-precipitation correlations were seen in summer mainly over northern low-temperature areas of Siberia and over arid areas to the south of 60^oN, respectively. Since the highest ET-NDVI correlation was found in June when the vegetation in the area is most active, it is considered that the primary factor for the ET interannual change should be vegetation interannual change. However, the strong correlations in ET-temperature and ET-precipitation suggest that these two climate factors have a significant relation to the ET interannual change in addition to the vegetation. This result will be a lead to reveal the relationship between the climate-vegetation system and its influence on the ET.