Spatio-temporal variation in surface water in Punjab, Pakistan from 1985 to 2020 using machine-learning methods with time-series remote sensing data and driving factors

Pakistan is home to many natural and artificial bodies of water, which are inevitable for agriculture, domestic use, recreation, etc. In the arid, semi-arid, and wet areas of the land, the distribution, spatio-temporal variations, and the impacts on water dynamics of climate and anthropogenic drivers were studied. In this study, we used Landsat, 5, 7, and 8 satellite images to generate annual and seasonal water frequency maps per pixel for the 1985-2020 period. The analysis findings have shown substantial inter-and intra-annual variability in rainfall and temperature. In addition, results revealed significant zonal disparities in water patterns, with the arid zone displaying the most drastic variations. We found out that the area (2530.42 km2) is occupied by water bodies, of which 1322.24 km2 (52.25%) seasonal and 1208.18 km2 (47.75%) are permanent water zones. There is a dramatic decline rate of 1.02 +/- 1.84 km2/year in contrast to permanent water (0.97 +/- 1.99 km2/year), total inland seasonal water has increased. During July and August, Punjab has the highest seasonal water area (1822.30 km2) and declines to the lowest (523.20 km2) in October to December and February to May when permanent water (708.12 km2) is greater than that of seasonal water. Gross Domestic Product (GDP) and rainfall were positively related to the surface water areas, while the temperature was inversely related. The outcome of our research offers valuable insights into future spatio-temporal variations, the supply of surface water in Punjab in the context of such anthropogenic and climate change activities.

Automated summary

This study investigated the distribution, spatio-temporal variations, and impacts on water dynamics of climate and anthropogenic drivers in Pakistan using satellite images. The findings revealed significant variability in rainfall and temperature, as well as disparities in water patterns, with the arid zone experiencing the most drastic variations. The study also identified a significant decline in water bodies area over the study period, with GDP and rainfall positively related to surface water areas.