Applicability of Multispectral Images to Detect Soil Organic Carbon Content in Land Suitability Assessment: A Case of a Sugarcane Plantation

Abstract

Soil organic carbon is important for sugarcane production as it plays a significant role in the development of soil aggregates and soil improvement in any agricultural soil, hence increasing soil health. Increasing soil organic matter encourages soil aggregation and slows the rate of organic matter breakdown. Soil aggregates act as nuclei for soil stabilization with time. It improves soil attributes by increasing organic carbon content in the soil and enhances the physical properties of the soil that favor water infiltration and retention. However, traditional methods of SOC determination like laboratory analyses are expensive and time-consuming. The objective of the study was to assess the applicability of multispectral images to determine SOC content in the surface soil. The research adopted two means to determine SOC of sugarcane. Under the first method, the lands were selected according to the sugarcane productivity in the study area, namely: low (35-54 t ha-1), medium (55-79 t ha-1), and high (80-100 t ha-1) productivity. Soil samples were collected up to 15cm depth. Walkley and Black Method was used for SOC determination in the laboratory. Simultaneously, multispectral images of each land were obtained using a drone platform. Multispectral images were then used to calculate Normalized Difference Vegetation Index (NDVI), Bare Soil Index (BSI), and Modified Secondary Soil Adjusted Vegetation Index (MSAVI2). These indices were used to predict SOC. The outcome was compared with SOC obtained from the laboratory analysis results. The results showed that the soil organic carbon (SOC) varied between 2.72 and 3.63% from the mean for the   the highest and lowest productivity lands respectively. The results of regression analysis observed a moderate correlation between SOC and the BSI values (0.4828). Weak correlations were observed in MSAVI2 (0.0269) and NDVI (0.0858). Future research should focus on improving these indices for SOC determination with increasing sample quantity.