Get More Info!

Announcement
Announcement

ANNOUNCEMENTS

  • Applications for PhD admissions are open
  • Announcing admissions for the academic session commencing July 2020 for all programmes - Apply Now
Characterization of terrain and vegetation indices in assessing soil erosion risk

Student name: Ms Khushboo Jain
Guide: Dr Chander Kumar Singh
Year of completion: 2018
Host Organisation: Indian Institute of Remote Sensing, Dehradun
Supervisor (Host Organisation): Mr Justin George
Abstract:

Soil erosion is a major cause for soil degradation and loss of upper fertile layer of the earth for agricultural purposes especially on mountainous topography. The following study is an innovative and practical effort to integrate various relative vegetation and terrain indices for soil erosion studies in an area. It offers an easy and creative approach with reduced time and cost of required data for creating erosion risk maps of a specified area .Vegetation cover is a major factor which affects soil degradation and conservation in an area. RUSLE model is a combined study of six soil erosion factors in which c factor is most often very hard to predict. This study was done on Tehri Garhwal district in the north western part of Uttarakhand using Landsat 8 30m resolution data of year 2014-2017. Random 70 sample points were marked on the study area of which C factor values were extracted using the processed NDVI image. The values of the C factor were then plotted using five different regression equations developed by works of various authors and then compared using model diagnostics and information criterion statistics. Knijff exponential model developed for temperate region performed the worst and Durigon’s linear model for tropical regions was found to be best suited for our study area.

Soil erosion risk map was prepared by using three major terrain indices derived using SRTM Dem data of 1 arc second that is 30 m resolution; topographic wetness index(TWI), length and slope factor(LS factor) and Stream power index(SPI) which were combined with pre-processed NDVI, SAVI and MSAVI maps of pre and post monsoon seasons. Multivariate classification using Isodata clustering and maximum likelihood algorithim was used to classify the study area into four potential soil erosion classes. Results showed that a clear increase in soil erosion can be seen in post monsoon season especially in areas with steeper slopes and lower vegetation cover in the studied area. RUSLE model was used to produce values of soil erosion classes using a combination of various factors(R,K,LS,C,P).The annual erosion rate was estimated to be around 126 ton per hectare in post monsoon season. Thus the soil erosion maps produced using RUSLE model can help strategists to develop mitigation strategies especially for areas prone to higher erosion values.

Keywords:-Flaash, Terrain indices, vegetation indices, Regression models, RUSLE model