Monitoring Water Availability in Northern Inland Waters from Space

Sammanfattning: River deltas and lakes support biodiversity and offer crucial ecosystem services such as freshwater provision, flood control, and fishing. However, climate change and human activities have affected deltas and lakes globally, altering the services they provide. Since delta and lake surface water occurrence and water levels respond to climate change and anthropogenic activities, we need to monitor their variations to understand the potential drivers for effective water management strategies. However, important deltas like the Selenga River Delta (SRD) in Russia lack a detailed analysis of water occurrence. Regarding lake water level, there has been a decline in the number of gauging stations globally, due to installation and maintenance costs. For example, Sweden has ~100,000 lakes which are sources of freshwater and hydro-power, but only 38 lakes have long and continuous in-situ records of water level.As satellite data are reliable alternatives for conventional methods to monitor deltas and lakes, I employed Earth Observations (EO) to quantify changes in surface water occurrence in the SRD and water levels in Swedish lakes and identify their main drivers. I also developed and explored a novel methodology for lake water level estimation based on Differential Interferometric Synthetic Aperture Radar (D-InSAR) by calculating the six-day phase differences in 30 Swedish lakes.To achieve these objectives, I trained and applied a Maximum Likelihood classification to Landsat images from 1987 to 2020 and quantified surface water occurrence and its changes in the SRD. I found that surface water occurrence in 51% of the delta experienced a decrease. As the Selenga River is the only river flowing into the SRD, the change in surface water occurrence in the SRD correlated with river discharge, but not with the river suspended sediment concentration, the lake water level in the outlet of the SRD, or evapotranspiration over the delta.In Sweden, I used satellite altimetry data from ERS-2, ENVISAT, JASON-1,2,3, SARAL, and Sentinel-3A/B to quantify water levels in 144 lakes from 1995-2022. I found that 52% of the lakes showed increasing trends (mostly in the north) and 43% decreasing trends (mostly in the south). Water level trends and variabilities did not correlate strongly with hydroclimatic changes (precipitation and temperature) but differed in regulated lakes compared to unregulated ones, both in the north and in the south of Sweden.The results of the D-InSAR method for water level estimation in two Swedish lakes (Hjälmaren and Solnen) showed that with water level changes smaller than a complete SAR phase, the phase changes correlate with in-situ water level changes with a minimum Root Mean Square Error of 0.43 cm in some pixels. In all 30 lakes, I accumulated the phase changes of each pixel throughout the whole number of interferograms to construct water levels. This method replicated the direction of water level changes shown by high Pearson’s correlations in at least one pixel in each lake.This thesis highlights the importance of EO for estimating surface water occurrence and lake water levels and brings focus to the future of EO through advanced space missions such as Surface Water and Ocean Topography (SWOT) and NASA-ISRO Synthetic Aperture Radar (NISAR). The findings underscore the need to continuously monitor lake water level and occurrence to adapt to climate change and understand the effects of water-regulatory schemes.