EXTRAPOLATION OF POINT MEASUREMENTS OF EVAPORATION - SOME ISSUES OF SCALE

The paper focuses on extrapolation of observed values as a means to determine regional evaporation. Evaporation data from diverse plant communities in southern Australia were drawn together to assess the magnitude and causes of areal variation in the process over the landscape. Soil water effects on daily evaporation rates were responsible for pronounced variability over a few hectares of uniform vegetation, variability being comparable to that encountered at an extended scale subject to combined influences of soil water and vegetation effects. In the longer term, local effects of soil water on evaporation were apparent, albeit with attenuated areal variability. Short term differences between extensive plant communities did not necessarily persist, sometimes resulting in a reversal of differences in evaporation rate. Estimating regional evaporation at time scales ranging from daily to monthly calls for an understanding of spatial and temporal variation in factors imposing control at the surface, especially those dealing with biological response. Improving this understanding to achive accuracy of estimates means that demands for finer temporal resolution in descriptions of regional evaporation must be accompanied by greater density of measurement points to resolve areal variability in flux rate. Further complexities were identified with advective effects caused by discontinuities in the landscape, notably for irrigated regions. Accounting for an advective enhancement of 20% in evaporation rate in an irrigation region involved a description of recurrent developing boundary layers at downwind scales of several hundreds of metres. Scale considerations for the developing characteristic boundary layer were also an issue in extrapolating data from a ventilated chamber study to predict evaporation response in a future environment enriched in atmospheric carbon dioxide. A potential change in evaporation rate by 50% accompanying the doubling of atmospheric carbon dioxide in a ventilated chamber was suggested as being moderated to a 10-20% reduction at the extended scale.