Research On Spatial-temporal Variation Characteristics And Its Causes Of Global Evaporation Based On Multi-reanalysis Datasets

Evaporation plays a crucial role in the global hydrological and energy cycle with implications ranging from water management to weather forecast and climate change assessments. Moreover, evaporation as the main way of water consumption is one of the most active processes affected by many environmental factors. Changes in the rate of evaporation over time are of considerable interest, especially in a warming climate. However, the lack of direct observations hampers the efforts to identify and quantify the change of the global evaporation. Moreover, the hydrological cycle is among the most uncertain aspects of climate model predictions. As a result, there remains considerable uncertainty concerning the magnitude of tropical oceanic evaporation response to a given increase in temperature. For now, the fundamental question of whether evaporation is increasing or decreasing, and, if so, where and when this is occurring, has not yet been comprehensively answered. The availability of the atmospheric reanalysis in recent years has provided a new opportunity to study not only the climatology, but also the variability of evaporation at different timescales. Therefore, we extracted the main characteristics of spatial-temporal changes in global evaporation based on multi-reanalysis datasets. Meanwhile, reanalyses are evaluated by comparing quantities with each other. Differences in the total amount, spatial variability, and distribution of evaporation are analyzed in order to estimate the uncertainties incorporated in these datasets. We can get these conclusions:(1) Long-term mean annual evaporation obtained from different reanalyses are consistent over most regions, with significant maritime-continental contrasts, as well as differences in meridional directions, and the land actual evapotranspiration generally decreases with the increase of altitude. In general, MERRA and NCEP-R2 may appropriately reflect the spatial-temporal characteristics of global evaporation, show strong representativeness. The CFSR and ERA-40 are capable of revealing the main characteristics of land actual evapotranspiration, whereas ERA-Interim, NCEP-R1, OAFlux, and HOAPS are relatively applicable for research focused on the evaporation over the oceans. According to ERA-40, NCEP-R1, and OAFlux, global evaporation significantly decreased for the period of 1958–1978. In contrast, most of the eight reanalyses show a significant linear increase trend for the period of 1979–2011, and evaporation over the oceans were even more pronounced.(2) For China, actual evapotranspiration decreases from southeast coastal to northwest inland regions, although differences among the reanalyses are apparent in Northwest China, northwestern Qinghai-Tibet Plateau, and southeast coastal area. Interannual fluctuations of actual evapotranspiration over China are distinct among reanalyses. Notably, NCEP-R2 and JRA-55 are exactly similar, both show significant increase trend during 1979~2013, while MERRA agreed well with ERA-Interim; however, NCEP-R1 reveals large differences with other reanalysis. Actual evapotranspiration mainly concentrated in summer, accounting for about 43.0% of the total actual evapotranspiration, evident seasonal variation. Our analysis supports the interpretation that the relationship between pan evaporation and actual evapotranspiration was complementary with water control and proportional with energy control.(3) By subdividing global land into nine climatic regions according to the Aridity Index(AI) and the mean annual potential evapotranspiration, we found that actual evapotranspiration increased significantly in humid regions from 1979 to 2013, while actual evapotranspiration increased significantly in the twenty-first century in non-humid regions. In the non-humid regions, precipitation is an important factor affecting the change of actual evapotranspiration. Conversely, actual evapotranspiration is more sensitive to the change of potential evapotranspiration in humid regions. In view of data availability, reanalysis datasets are analyzed and compared with actual evapotranspiration estimates calculated using the Budyko equation. Overall, NCEP-R2 shows the highest skill, MERRA comes second. ERA-Interim and JRA-55 have substantial problems on describing the variability of actual evapotranspiration, especially in the extreme humid region.(4) Tropical oceanic evaporation shows an interdecadal variation with a linearly increasing trend from 1979 to 2001 and remains more or less steady afterwards. From 1979 to 2013, dominant modes of tropical oceanic evaporation among the six reanalyses are extracted using multivariate empirical orthogonal function analysis(MV-EOF). Accordingly, the first MV-EOF mode represents the enhancement of evaporation over most of the tropical oceans from 1979 to the current period. The next three MV-EOF modes represent the interannual variability of tropical oceanic evaporation on time scales of El Ni?o–Southern Oscillation(ENSO), and evaporation anomalies associated with different types of ENSO are distinct. The present study proposes that sea surface temperature(SST) can impact evaporation by two ways. The most common way is that surface air temperature(SAT) trends largely reflect the state of SST. Evaporation is expected to intensify in a warmer climate, because warmer air can hold more water vapour. Indirectly, positive SST anomaly can stimulate anomalous low sea level pressure(SLP) centred at the warming environment. It has a positive contribution to the water vapour convergence, which will inhibit evaporation. Simultaneous increase in SAT and atmospheric water vapour content exerts an opposing effect on atmospheric evaporative demand. Large-scale changes in tropical oceanic evaporation are mainly attributed to this thermodynamic feedback.