The effect of choosing three different C factor formulae derived from NDVI on a fully raster-based erosion modelling

2nd International Conference of Indonesian Society for Remote Sensing 2016

地球科学;计算机科学

Sulistyo, Bambang^1

Department of Soil Science, Faculty of Agriculture, University of Bengkulu Jalan WR Supratman, Kandang Limun, Bengkulu

38371, Indonesia^1

关键词: Correlation value;    Digital elevation model;    Erosion modelling;    Land-cover classification;    Model validation;    Remote sensing data;    Research results;    Threshold-value;   

Others  :  https://iopscience.iop.org/article/10.1088/1755-1315/47/1/012030/pdf
DOI  :  10.1088/1755-1315/47/1/012030

学科分类：计算机科学（综合）

来源: IOP

【 摘 要 】

The research was aimed at studying the efect of choosing three different C factor formulae derived from NDVI on a fully raster-based erosion modelling of The USLE using remote sensing data and GIS technique. Methods applied was by analysing all factors affecting erosion such that all data were in the form of raster. Those data were R, K, LS, C and P factors. Monthly R factor was evaluated based on formula developed by Abdurachman. K factor was determined using modified formula used by Ministry of Forestry based on soil samples taken in the field. LS factor was derived from Digital Elevation Model. Three C factors used were all derived from NDVI and developed by Suriyaprasit (non-linear) and by Sulistyo (linear and non-linear). P factor was derived from the combination between slope data and landcover classification interpreted from Landsat 7 ETM+. Another analysis was the creation of map of Bulk Density used to convert erosion unit. To know the model accuracy, model validation was done by applying statistical analysis and by comparing Emodelwith Eactual. A threshold value of ≥ 0.80 or ≥ 80% was chosen to justify. The research result showed that all Emodelusing three formulae of C factors have coeeficient of correlation value of > 0.8. The results of analysis of variance showed that there was significantly difference between Emodeland Eactualwhen using C factor formula developed by Suriyaprasit and Sulistyo (non-linear). Among the three formulae, only Emodelusing C factor formula developed by Sulistyo (linear) reached the accuracy of 81.13% while the other only 56.02% as developed by Sulistyo (nonlinear) and 4.70% as developed by Suriyaprasit, respectively.