Large scale estimation of evapotranspiration

Li, Fuqin (1999) Large scale estimation of evapotranspiration. PhD thesis, Murdoch University.

Abstract

Evapotranspiration is an essential component of the energy and water bud­get, but its estimation depends on available data sources and the environment of an area. Remote sensing techniques, combined with routine meteorological data, are used to estimate evapotranspiration over central Australia through the development and application of a number of models, ranging from physically based instantaneous models to a daily simulation model.

The proposed models are evaluated using aircraft observations over two dis­tinct vegetation regimes in south-western Australia. Among the three physically based instantaneous models, single-source models using an excess resistance term empirically determined performed better than a two-source model which does not require such a parameterization. The mean absolute difference between measured and estimated values of the sensible heat flux is below 17 wm-2 in comparison to approximately 40 Wm-2 for evapotranspiration. Estimates of evapotranspiration depend on the closure of the surface energy balance and incorporate all residual errors in this closure. All models perform better over the agricultural vegetation than over the native vegetation.

As these physically based models only provide instantaneous estimates of evapotranspiration at satellite overpass, a coupled one dimensional soil-canopy­atmosphere model and a simple budget water balance model have been used to simulate the daily evapotranspiration. Comparison of these results with the air-craft observations shows that the coupled model provides a good estimate of sur­face heat fluxes over the agricultural area with mean absolute differences between measured and estimate values being approximately 25 wm-2 for both sensible heat flux and evapotranspiration. Over the native vegetation, the mean absolute difference between measured and observed fluxes increased to 49 and 47 wm-2, respectively, for the sensible heat and evapotranspiration. This increase results from the inability of a simple water balance model to incorporate the effects of the underlying aquifer on deep rooted native vegetation, particularly during the dry summer season. It also highlights the sensitivity of the one dimensional soil-canopy-atmosphere model to the specification of soil moisture.

Since the model simulation of surface temperature is also very sensitive to the soil moisture, a comparison between model simulation of surface temperature and satellite derived surface temperature was used to adjust parameters of a water balance model resulting in better estimates of soil moisture and consequently improved predictions of evapotranspiration.

These models have been applied to estimating evapotranspiration in central Australia, using limited routine meteorological data and the NOAA-14 AVHRR overpass. Minimizing the difference between model predicted surface temperature and satellite derived temperature to adjust the estimated soil moisture, both the instantaneous physically based model and the simulation yielded consistent re­sults for 8 representative clear sky days during 1996-1997. These results highlight the sensitivity of surface temperature to soil moisture and suggest that radiomet­ric surface temperature can be used to adjust simple water balance estimates of soil moisture providing a simple and effective means of estimating large scale evapotranspiration in remote arid regions.

Item Type:	Thesis (PhD)
Murdoch Affiliation(s):	School of Environmental Science
Notes:	Note to the author: If you would like to make your thesis openly available on Murdoch University Library's Research Repository, please contact: repository@murdoch.edu.au. Thank you.
Supervisor(s):	Lyons, Tom
URI:	http://researchrepository.murdoch.edu.au/id/eprint/51652

Item Control Page