Abstract Agricultural land systems are highly susceptible to climatic variability, particularly in semi-arid regions where declining and erratic rainfall patterns threaten land cover stability and water security. While land use/land cover (LULC) dynamics have been widely studied, limited attention has been paid to integrating long-term rainfall trends with pixel-level LULC transitions to interpret land degradation and hydrologic vulnerability. This study evaluates climate-induced transformations in Khammam district by integrating 30 years of satellite-derived LULC maps (1990–2020) with historical rainfall records using transition matrices, net change analysis, and Sankey flow visualizations. The overlay of rainfall trends with LULC dynamics revealed a significant contraction in agricultural land – from 3400 km2 in 1990 to 2700 km2 in 2020 – coinciding with a concurrent decline in monsoonal rainfall of over 150 mm. Transition matrix analysis showed that over 700 km2 of cropland was converted to fallow or barren land, while forest cover declined modestly and built-up areas expanded across peri-urban zones. A Granger causality test was performed to statistically validate the influence of rainfall on agricultural transitions. Results confirmed a significant causal relationship at lag-1 and lag-2 years (F-statistics of 4.31 and 3.87; p-values of 0.038 and 0.041, respectively), reinforcing that rainfall variability precedes agricultural area shrinkage. The findings suggest a cascading impact: rainfall decline → agricultural degradation → groundwater pressure and eventual land abandonment. This highlights the urgent need to integrate climate-sensitive indicators into land management strategies and water governance policies. The study advocates for spatially explicit early warning systems, vulnerability indices, and adaptive planning frameworks that enhance land and water resource resilience under projected climate extremes. Research highlights <UnorderedList Mark="Bullet"> <ItemContent> This study combines 30 years (1990–2020) of satellite-derived land use/land cover (LULC) maps with gridded and station-validated rainfall datasets to look at how climate affects land use and water resources in a semi-arid Indian district. </ItemContent> <ItemContent> Quantification of climate-induced agricultural decline: Agricultural land diminished by approximately 700 km2 over three decades, paralleling a statistically significant reduction in monsoonal rainfall (approximately 3.5 mm yr⁻1), signifying the pronounced climate sensitivity of rain-fed agricultural systems. </ItemContent> <ItemContent> Granger causality analysis shows that changes in rainfall come before changes in agricultural land by 1 to 2 years (p < 0.05), making rainfall decline an early warning sign of land degradation. </ItemContent> <ItemContent> Spatially explicit visualization of land conversion pathways: Transition matrices and Sankey flow diagrams show that the most common changes from agriculture to fallow and from agriculture to built-up land are happening. This shows how climate stress and urban growth can have a ripple effect on the stability of the landscape. </ItemContent> <ItemContent> Policy-relevant sustainability assessment: A PCA-based sustainability index finds peri-urban and former agricultural hotspots to be very vulnerable areas. This information can be used to plan land and water resources in semi-arid areas that can adapt to climate change. </ItemContent> </UnorderedList>

Spatio-temporal integration of rainfall trends and land use change for water resource assessment

期刊论文
发表日期：

2026年2月23日

查看全文

Shiphali Preeti Aind,
A R Quaff

摘要

Abstract

Agricultural land systems are highly susceptible to climatic variability, particularly in semi-arid regions where declining and erratic rainfall patterns threaten land cover stability and water security. While land use/land cover (LULC) dynamics have been widely studied, limited attention has been paid to integrating long-term rainfall trends with pixel-level LULC transitions to interpret land degradation and hydrologic vulnerability. This study evaluates climate-induced transformations in Khammam district by integrating 30 years of satellite-derived LULC maps (1990–2020) with historical rainfall records using transition matrices, net change analysis, and Sankey flow visualizations. The overlay of rainfall trends with LULC dynamics revealed a significant contraction in agricultural land – from 3400 km² in 1990 to 2700 km² in 2020 – coinciding with a concurrent decline in monsoonal rainfall of over 150 mm. Transition matrix analysis showed that over 700 km² of cropland was converted to fallow or barren land, while forest cover declined modestly and built-up areas expanded across peri-urban zones. A Granger causality test was performed to statistically validate the influence of rainfall on agricultural transitions. Results confirmed a significant causal relationship at lag-1 and lag-2 years (F-statistics of 4.31 and 3.87; p-values of 0.038 and 0.041, respectively), reinforcing that rainfall variability precedes agricultural area shrinkage. The findings suggest a cascading impact: rainfall decline → agricultural degradation → groundwater pressure and eventual land abandonment. This highlights the urgent need to integrate climate-sensitive indicators into land management strategies and water governance policies. The study advocates for spatially explicit early warning systems, vulnerability indices, and adaptive planning frameworks that enhance land and water resource resilience under projected climate extremes.

Research highlights

This study combines 30 years (1990–2020) of satellite-derived land use/land cover (LULC) maps with gridded and station-validated rainfall datasets to look at how climate affects land use and water resources in a semi-arid Indian district.

Quantification of climate-induced agricultural decline: Agricultural land diminished by approximately 700 km² over three decades, paralleling a statistically significant reduction in monsoonal rainfall (approximately 3.5 mm yr⁻¹), signifying the pronounced climate sensitivity of rain-fed agricultural systems.

Granger causality analysis shows that changes in rainfall come before changes in agricultural land by 1 to 2 years (p < 0.05), making rainfall decline an early warning sign of land degradation.

Spatially explicit visualization of land conversion pathways: Transition matrices and Sankey flow diagrams show that the most common changes from agriculture to fallow and from agriculture to built-up land are happening. This shows how climate stress and urban growth can have a ripple effect on the stability of the landscape.

Policy-relevant sustainability assessment: A PCA-based sustainability index finds peri-urban and former agricultural hotspots to be very vulnerable areas. This information can be used to plan land and water resources in semi-arid areas that can adapt to climate change.