Improving Land Use/Cover Classification Accuracy from Random Forest Feature Importance Selection Based on Synergistic Use of Sentinel Data and Digital Elevation Model in Agriculturally Dominated Landscape

Abstract

Land use and land cover (LULC) mapping can be of great help in changing land use decisions, but accurate mapping of LULC categories is challenging, especially in semi-arid areas with extensive farming systems and seasonal vegetation phenology. Machine learning algorithms are now widely used for LULC mapping because they provide analytical capabilities for LULC classification. However, the use of machine learning algorithms to improve classification performance is still being explored. The objective of this study is to investigate how to improve the performance of LULC models to reduce prediction errors. To address this question, the study applied a Random Forest (RF) based feature selection approach using Sentinel-1, -2, and Shuttle Radar Topographic Mission (SRTM) data. Results from RF show that the Sentinel-2 data only achieved an out-of-bag overall accuracy of 84.2%, while the Sentinel-1 and SRTM data achieved 83% and 76.44%, respectively. Classification accuracy improved to 89.1% when Sentinel-2, Sentinel-1 backscatter, and SRTM data were combined. This represents a 4.9% improvement in overall accuracy compared to Sentinel-2 alone and a 6.1% and 12.66% improvement compared to Sentinel-1 and SRTM data, respectively. Further independent validation, based on equally sized stratified random samples, consistently found a 5.3% difference between the Sentinel-2 and the combined datasets. This study demonstrates the importance of the synergy between optical, radar, and elevation data in improving the accuracy of LULC maps. In principle, the LULC maps produced in this study could help decision-makers in a wide range of spatial planning applications.

Performance Evaluation of DL-based Models for LULC segmentation using SAR time series data of regions in Gujarat

“…Starting with the field survey, Adrian et al, [63] labelled thirteen crops, including urban, water, and soil, with the help of field data and WorldView-3 (WV3) images; Wang et al, [64] also conducted the field survey; Solórzano et al, [50] created ten land cover features by using Sentinel-2 images, Planet images, VHR (Very High Resolution) images from google earth and also conducted the field survey. Priyatna et al, [65] labelled ten classes with the help of Landsat-8 image, and Ibrahim et al, [1] created eight land cover features by using RGB Composite images and google earth images.The authors of the papers [25,61,66,67] and [68] did not specify the source of their ground-truth data and did not go into great detail about land cover aspects.…”

“…Since the SAR image has a reasonable spatial resolution, land cover features can be improved and increased for this investigation, even though the model performed well in its application.Table 2 also discussed the Deep Learning or Machine learning techniques applied to SAR and Optical images. The authors of the papers [1,25,32] integrated SAR and Optical images for land cover classification using the Random Forest algorithm. Song et al, [28] used U-Net for classifying seven land cover features.…”

“…Combating land degradation and regenerating degraded land and soil, particularly in areas impacted by desertification, drought, and floods, will be a top priority by 2030. [1] Over the last few decades, the increase in land use and land cover changes (LULC) in natural ecosystems has harmed carbon stocks, climate change, biodiversity, and the global climate [2][3][4][5]. Deforestation caused by urbanisation and agricultural expansion is regarded as one of the most serious environmental threats in the twenty-first century [6].…”

“…Because of their consistency and repeatability at local and large spatial scales, satellite Earth observation data are one of the most reliable options for monitoring land degradation in the context of the Sustainable Development Goals (SDGs). [1] Regarding spaceborne/airborne images, classification entails assigning classes/labels to each spatial unit or pixel. The type of human actions (such as buildings, bridges/highways, rail tracks, airports, dams, and reservoirs) and natural components (such as mountains, plateaus, hills, minerals, and forests) on the landscape over time using established scientific and statistical methods of analysis of significant materials is known as land cover mapping or land use land cover [7].…”

Multi-Year Cereal Crop Classification Model in a Semi-Arid Region Using Sentinel-2 and Landsat 7–8 Data

“…According to Yang et al [19], the random forest (RF) [20] algorithm has been widely utilized in previous LULC studies [16,19,[21][22][23][24][25][26][27][28][29], followed by support vector machine (SVM) [30][31][32][33][34][35]. Li et al [36] conducted a comparison between the performance of the RF classification algorithm and other algorithms such as SVM and artificial neural network (ANN).…”