Exploring the Potential of Soil Salinity Assessment through Remote Sensing and GIS: Case Study in the Coastal Rural Areas of Bangladesh

Soil salinity is a negative impact of climate change, and it is a significant problem for the coastal region of Bangladesh, which has been increasing in the last four decades. The issue of soil salinity substantially limits the agricultural crop production in coastal areas. Therefore, a soil salinity assessment is essential for proper land-use planning in agricultural crop production. This research was carried out to determine the soil salinity area with different salinity levels in Barguna Sadar Upazila (sub-district). The remote sensing technique, which is a potentially quick yet effective method for the soil salinity estimation in data-scarce conditions, was applied. The methodology employed the Landsat 8 OLI dataset along with nine soil salinity indices to develop a soil salinity map. The maps were from Soil Resource Development Institute (SRDI), and low NDVI value (-0.01 to 0.48) was produced using satellite images illustrate the extent of the soil salinity for the study area. However, nine linear regressions, which were made between the pixel value of the satellite-based generated map and ground truth soil salinity data, that is, the EC value, indicate a maximum R-2 value for the salinity index SI 7 = G x R/B, representing a value of 0.022. This minimal R-2 value indicates a negligible relationship between the ground EC value and the pixel value of the salinity index generated map, inferring that the indices are not sufficient to assess the soil salinity. Nonetheless, this research's findings offer a guide for researchers to investigate alternative geospatial approaches for this geophysical condition.

Automated summary

Soil salinity is a significant problem for the coastal region of Bangladesh, and this study aimed to assess the soil salinity using remote sensing techniques. The findings indicate that the current soil salinity indices are insufficient for accurate assessment. Alternative geospatial approaches should be explored.