Assessing the Soil Fertility Using Landsat TM Imagery and Geospatial Statistical Analysis
This paper aims to investigate the soil fertility of Shunyi District’s cropland combing remote sensing and ground census data based on the Geostatistical Analyst of ArcGIS. Firstly, Landsat TM image was used to identify the spatial distribution and estimate the cropland plot area using support vector machine (SVM) classification method, and the overall classification is 91.5 % by 435 field survey points. Then, the survey indicators were added to ArcGIS such as organic matter, available P, available K, total N, soil pH, etc. After exploring the sample data for each indicator, trend surfaces were generated using the optimum prediction models after cross validation. Finally, according to the identified cropland plots, the soil quality index (SQI) was derived to map the soil fertility of the study area. The result shows that the southwestern part and northeastern corner of this district were found to be high in soil pH, which lies in between 8.2 and 8.6. Additionally, wide variability of organic matter, total N, available P and K were noted which can be due to the extent of cultivation in these areas while the change in fertility level could be due to anthropogenic influence. When considering the soil heavy metals, Zn, Fe, Cu and Mn show almost the same distribution.
Weiguo Pan, Jianxing Ren and Yongguang Li
J. L. Zhao et al., "Assessing the Soil Fertility Using Landsat TM Imagery and Geospatial Statistical Analysis", Advanced Materials Research, Vols. 347-353, pp. 3559-3563, 2012