No abstract

A systematic method is proposed to optimally combine estimates of the terrestrial water budget from different data sources and to enforce the water balance constraint using data assimilation techniques. The method is applied to create global long-term records of the terrestrial water budget by merging a number of global datasets including in situ observations, remote sensing retrievals, land surface model simulations, and global reanalyses. The estimation process has three steps. First, a conventional analysis on the errors and biases in different data sources is conducted based on existing validation/error studies and other information such as sensor network density, model physics, and calibration procedures. Then, the data merging process combines different estimates so that biases and errors from different data sources can be compensated to the greatest extent and the merged estimates have the best possible confidence. Finally, water balance errors are resolved using the constrained Kalman filter technique. The procedure is applied to 32 globally distributed major basins for 1984-2006. The authors believe that the resulting global water budget estimates can be used as a baseline dataset for large-scale diagnostic studies, for example, integrated assessment of basin water resources, trend analysis and attribution, and climate change studies. The global scale of the analysis presents significant challenges in carrying out the error analysis for each water budget variable. For some variables (e.g., evapotranspiration) the assumptions underpinning the error analysis lack supporting quantitative analysis and, thus, may not hold for specific locations. Nevertheless, the merging and water balance constraining technique can be applied to many problems.

No abstract

International audienceEstimating evapotranspiration (ET) at continental to global scales is central to understanding the partitioning of energy and water at the earth's surface and the feedbacks with the atmosphere and biosphere, especially in the context of climate change. Recent evaluations of global estimates from remote sensing, upscaled observations, land surface models and atmospheric reanalyses indicate large uncertainty across the datasets of the order of 50% of the global annual mean value. In this paper, we explore the uncertainties in global land ET estimates using three process-based ET models and a set of remote sensing and observational based radiation and meteorological forcing datasets. Input forcings were obtained from International Satellite Cloud Climatology Project (ISCCP) and Surface Radiation Budget (SRB). The three process-based ET models are: a surface energy balance method (SEBS), a revised Penman-Monteith (PM) model, and a modified Priestley-Taylor model. Evaluations of the radiation products from ISCCP and SRB show large differences in the components of surface radiation, and temporal inconsistencies that relate to changes in satellite sensors and retrieval algorithms. In particular, step changes in the ISCCP surface temperature and humidity data lead to spurious increases in downward and upward longwave radiation that contributes to a step change in net radiation, and the ISCCP data are not used further. An ensemble of global estimates of land surface ET are generated at daily time scale and 0.5 degree spatial resolution for 1984-2007 using two SRB radiation products (SRB and SRBqc) and the three models. Uncertainty in ET from the models is much larger than the uncertainty from the radiation data. The largest uncertainties relative to the mean annual ET are in transition zones between dry and humid regions and monsoon regions. Comparisons with previous studies and an inferred estimate of ET from long-term inferred ET indicate that the ensemble mean value is reasonable, but generally biased high globally. Long-term changes over 1984-2007 indicate a slight increase over 1984-1998 and decline thereafter, although uncertainties in the forcing radiation data and lack of direct linkage with soil moisture limitations in the models prevents attribution of these changes