Flash Water Detection in Mining Areas Using Satellite Imagery and Meteorological Data

Detection of temporary surface water is a critical aspect of monitoring the impact of mining on the environment and the impacts of underground gas storage (UGS) sites in particular. Recent years have been very dynamic for the energy sector, as increasing energy demand and the progression of global warming require constant improvements and changes. The general trend is to move away from conventional energy sources. However, green fuels require large storage volumes, which can be provided by geological storage. It can also be used to trap CO₂ and thus contribute to the reduction of CO₂ emission into the atmosphere.

Injection and withdrawal of gases in underground magazines affect the environment in a number of ways. Pressure changes and corresponding stress induce seismic activity, surface subsidence, and uplift. Gas leakage poses a risk to soil, water and air contamination. Therefore, it is necessary to treat UGS sites together with their surroundings in a holistic manner considering all potential impacts. The study UGS site is located in northern Poland. Natural gas is stored in salt caverns. The terrain to the north of the facility is of particular interest, as it is agricultural land with an old drainage system. It has not been maintained for years and the water reoccurs periodically. Low elevation and short distance to the sea favour flash flooding, which might be reinforced by UGS induced surface movements. The purpose of the test study is to detect surface water with a remote sensing based approach and establish the correlation between rainfall and surface water dynamics.

Satellite remote sensing provides a valuable means of continuous and large-scale monitoring of surface water dynamics. Spectral bands and indicators enable the discrimination of various types of land cover and their changes, including the appearance of flash water. The proposed methodology involves time-series analysis of open satellite data (Sentinel-2), spatial statistics, and comparative analysis of selected indicators and spectral bands used for water detection. Additionally, daily precipitation data from a local meteorological station were integrated into the analysis to evaluate the accuracy of surface water detection. Regression analysis has been done to analyse the relationship between the accumulation of water and rainfall, and therefore assess whether the indicators tested are suitable. 

The test analysis covers the period August 2015 - December 2023. The area of interest is cultivated. Crop fields, after harvest and before the vegetation season, are mostly bare soil, which can be mistakenly interpreted using basic plant moisture indices. Therefore, more combinations were tested and verified with rainfall data. By incorporating meteorological data, we aim to establish a more comprehensive understanding of the temporal variability in the presence of surface water near the gas storage site. The findings of this study contribute to the development of a complex monitoring system at a UGS site.

The research has been carried out under the project acronym CLEAR (grant no WPN/4/67/CLEAR/2022) financed from the Polish National Centre for Research and Development.