Actual Evapotranspiration Estimates for Australia Inter-comparison and Evaluation

[摘要] One objective of the Water Information Research and Development Alliance (WIRADA) is to assist the Bureau of Meteorology develop an operational gridded actual evapotranspiration (AET) product. Their main prospective use is to support national water accounts and assessments; hence the AET products need to cover the nation at a spatial resolution of at least 0.05° grid-cell resolution and at least monthly resolution. Here we evaluate 16 AETproducts representing eight AET estimation methods for consideration in operational production. The algorithms cover a wide spectrum of dynamic hydrological models and usesof satellite observations. We evaluated the products through inter-comparison and evaluation against AET estimates from flux-tower measurements at 7 sites and water balances for 514 catchments. An assessment of the pros and cons of each of the methods can be found in Table 19 in Section10.Two dynamic models (AWRA and AWAP) and one satellite algorithm (CMRS) provided theAET estimates that agreed best with the observations. The key advantage of dynamic model AET estimates for dry land environments is that they are constrained by rainfall. However this becomes a disadvantage for lateral inflow receiving areas, such as floodplains, wetlands, water bodies, irrigation areas and groundwater dependent ecosystems. Under such circumstances the algorithms using remote sensing generally perform better.Based on the assessment it is recommended that:an interim daily gridded AET product be developed based on the AWRA model, or possibly a simple average between the AWRA and AWAP models, substituted by estimates from the CMRS method for inflow areas.Key requirements beyond existing Bureau infrastructure and operational data streams are (1) operational ingestion and processing of NASA MODIS reflectance data stream, and (2)development of a mask delineating lateral inflow receiving areas. Based on the work in this report, an automated framework for evaluating the ongoing performance of the productionsystem should be established utilising the improving water information data set being developed by the Bureau. Such a framework could also be used to support a parallel effort tomonitor progress in algorithm research and development.We recommend further research to focus on:1. Collating or collecting data and developing methods to evaluate and estimate AET from inflow receiving areas, humid forests, arid regions and salt lakes.2. Specifying errors in the selected products and investigating possibilities to produce improved AET estimates using ensemble blending techniques.3. Investigating the potential of land surface temperature to improve AET estimates.4. Developing method to assimilate the CMRS product (or its input data) into the AWRA system.

[发布日期] 2011-04-01 [发布机构] CSIRO

[效力级别] Meteorology [学科分类] 地球科学(综合)