Estimation of the cover and management factor based on stratified coverage and remote sensing indices: a case study in the Loess Plateau of China

Abstract

We attempt to describe the cover and management (C) factor more comprehensively through the use of a simple and efficient method. We measure the coverage of each vegetation layer and C factor for 152 sampled plots in the Ansai watershed. We propose four stratified coverage indices (green coverage (V G), total coverage (V T), probability coverage (V P), weight coverage (V W)), derive green and yellow vegetation indices from Landsat 8 OLI images to reflect green and residue cover, and construct and validate C factor estimation models from stratified coverage and remote sensing indices, respectively. (1) V T and V P present C factor estimation advantages for grassland and shrub land. V W can better illustrate the C factor due to the relatively complete spatial structuring of woodland and orchard land. For cropland, four stratified coverage indices present the same estimation capacities for the C factor. Except for cropland and grassland, the estimation capabilities of V G are relatively low because the residue layer is ignored. (2) The C factor is more sensitive to yellow vegetation indices, which indicates that senescent fractional cover and litter are important and cannot be ignored. The linear and non-linear models can explain 56.6 and 61.8% of C factor variation, respectively, and the linear model is more accurate than the non-linear model. (3) Compared to traditional indices (projective coverage and single remote sensing indices), stratified coverage indices and a combination of several remote sensing indices can estimate the C factor more effectively. At the field scale, the C value estimation model can be selected according to the land-use type. At the watershed and regional scales, a linear model is recommended for C factor estimation.