COMPARISON OF 4 DIFFERENT STOMATAL-RESISTANCE SCHEMES USING FIFE OBSERVATIONS

Stomatal resistance (R-s) calculation has a major impact on the surfac e energy partitioning that influences diverse boundary layer processes . Present operational limited area or mesoscale models have the Jarvis -type parameterization, whereas the microscale and the climate simulat ion models prefer physiological schemes for estimating R-s. The pivota l question regarding operational mesoscale models is whether an iterat ive physiological scheme needs to be adopted ahead of the analytical J arvis-type formulation. This question is addressed by comparing the ab ility of three physiological schemes along with a typical Jarvis-type scheme for predicting R-s using observations made during FIFE. The dat a used is typical of a C4-type vegetation, predominant in regions of h igh convective activity such as the semiarid Tropics and the southern United States grasslands. Data from three different intensive held cam paigns are analyzed to account for vegetation and hydrological diversi ty. It is found that the Jarvis-type approach has low variance in the outcome due to a poor feedback for the ambient changes. The physiologi cal models, on the other hand, are found to be quite responsive to the external environment. All three physiological schemes have a similar performance qualitatively, which suggests that the vapor pressure defi cit approach or the relative humidity descriptor. used in the physiolo gical schemes may not yield different results for routine meteorologic al applications. For the data considered, the physiological schemes ha d a consistently better performance compared to the Jarvis-type scheme in predicting R-s outcome. All four schemes can, however, provide a r easonable estimate of the ensemble mean of the samples considered. A s ignificant influence of the seasonal change in the minimum R-s in the Jarvis-type scheme was also noticed, which suggests the use of nitroge n-based information for improving the performance of the Jarvis-type s cheme. A possible interactive influence of soil moisture on the capabi lities of the four schemes is also discussed. Overall, the physiologic al schemes performed better under higher moisture availability.