Spatial distribution of soil nutrients in a watershed of Himalayan landscape using terrain attributes and geostatistical methods

Abstract

Terrain attributes derived from digital terrain model (DTM) were used to study spatial variation of total soil C, N and available P in surface soils of a watershed of Himalayan landscape. Terrain attributes elevation, slope gradient and upslope catchment area (UCA) and terrain indices [terrain wetness index (TWI), water power index (WPI) and sediment transport index (STI)] were derived from DTM and evaluated for their potential in soil nutrients mapping. These nutrients showed positive correlation with UCA, TWI, SPI and STP terrain indices. Among these terrain indices, TWI showed highest correlation coefficient for TC (r 2 = 0.71), N (r 2 = 0.67) and P (r 2 = 0.66) followed by WPI and STI. Geostatistical analyses used to map these nutrients, co-kriging with TWI + NDVI, TWI and slope as co-variables, had improved the spatial prediction to 60.46, 55.81, 44.18 % for TC and 33.63, 21.78, 17.82 % for N, respectively, contrary to ordinary kriging. The prediction accuracy for P was improved with co-variables of TWI + NDVI and TWI by 30.03 and 4.50 %, respectively. The study clearly revealed that by integrating NDVI as co-variable has significantly improved the accuracy for TC followed by N and P. TWI alone as co-variable has improved the spatial prediction significantly.