Evaluation of Pedotransfer functions to estimate soil water limits

Abstract

Eight different pedotransfer functions (PTFs) namely Baumer, Brakensiek. British Soil Survey method (top soil), British Soil Survey method (sub soil), EPIC, Hutson, Manrique and Rawls developed for different regions were evaluated for their performance to estimate field capacity (FC) and permanent wilting point (;WP) of clay soils of Panubali watershed in Nagpur, Maharashtra. The input data requirement of these PTFs ranged from minimum textural data to additional bulk density and organic carbon. In correspondence tb measured data, the PTFs in general, tended to over-predict FC/PWP. Estimates of FC were relatively closer to measured values as compared to PWP, which were far off from the measured values. FC predicted by Rawls PTF / relatively accurate with the highest coefficient of determination (R2= 0.67) but root mean square error (RMSE) was> (0.05 m'm·)amongst the estimates of PWP, values predicted by Manrique PTF had lowest RMSE (0.04 m3m·') but R2 was 0.42, indicating / :;one to one correspondence between measured and predicted values. Except British Soil Survey method PTF for FC and PTF for PWP, estimates had higher RMSE, well above an average value (0.05 m'm·3) reported in the literature. Use of input data did not improve PTF performance as indicated by higher RMSE in estimates obtained by using PTFs requiring input. Study indicated that region specific PTF need to be developed .