Assessing the changes in hydrological budget including anthropogenic effect estimated by the reanalyses and rain-gauge based precipitation data over Asia

It has been pointed out that change in regional scale (100-1000 km) of land/water use induces the larger-scale climate changes such as desertification throughout the land-atmosphere (feedback) processes. However, few researches have been made for assessing the change of the hydrological budget from the atmospheric side. With the benefit of the reanalysis datasets, the 25-50 years change in atmospheric hydrological budget should be clarified, and increase in evapotranspiration (Et) caused by the increase in agricultural water use may be identified.

However, to assess evapotranspiration (Et) by the residual of precipitation (P) minus convergence (C), an accurate grid precipitation dataset is required in addition to the fine reanalyses datasets. Furthermore, it is desirable to prepare the precipitation grid data on the same interval as the reanalysis dataset.

Hence, an analysis has been done on the interannual change of hydrological budget over Asia, by using meteorological reanalyses data and a rain-gauge based grid precipitation dataset developed by the APHRODITE project (Yatagai et al., 2009), in which more than 4 times rain-gauge data available through the global telecommunication system (GTS) network is used and orographic precipitation is taken into account. As for the reanalysis, we used ERA40 data compiled by the European Centre for Medium Range Weather Forecast (ECMWF) and JRA25 compiled by Japan Meteological Agency. The study period is 1961-2002 with ERA40 and 1979-2004 with JRA25.

As a result, increase in evapotranspiration (Et) estimated as a residual between moisture convergence and precipitation is found at the semi-arid regions of China, Central Asia, Turkey and India etc, where the progress of desertification is reported. For some of these regions, these facts derived from the reanalysis and precipitation are consistent with the local reports that agricultural water use (irrigation by using river water, ground water, dam water) has been increased.

An attempt has been made to distinguish the anthropogenic increase of Et from natural variability by using the forecast evaporation (Ef) of reanalysis datasets. In the forecast phase of the four-dimensional data assimilation, atmospheric observation data and part of land surface observation data are assimilated at each assimilation time. On the contrary, definition of land use type, such as crop land, paddy field, forest, bare soil etc, are the same throughout the analysis years, although seasonal change of the parameters (albedo, roughness etc) may change.

We compared the trends between Et and Ef for the regions where the increase in Et were detected. As a result, Ef did not increase in most of these regions. Hence, it is speculated that the agricultural water use increases the evapotranspiration in summertime over the semi-arid regions. The ratio of the anthropogenic effects and horizontal distribution will be shown at the conference.