Spatiotemporal analysis of atmospheric methane over Kazakhstan using Sentinel-5P satellite observations between 2019–2024
Abstract
Methane (CH 4 ) is a potent greenhouse gas whose short-term radiative forcing significantly exceeds that of CO 2 , poses significant risks for regions with intensive hydrocarbon production, such as Kazakhstan.This study provides a country-wide spatiotemporal assessment of methane concentrations over Kazakhstan for 2019–2024 using Sentinel-5P TROPOMI data processed in Google Earth Engine. The analysis integrates statistical anomaly detection (98th and 99th percentile, and interquartile range methods) with Getis–Ord Gi* hotspot analysis and GIS-based regional quantification. The results demonstrate a clear upward trend in background methane levels, with December maxima increasing from 1881 ppb in 2019 to 1931 ppb in 2024, and a statistically significant annual increase of 8.23 ppb year -1 (Mann–Kendall p = 0.0085; 95% CI: 6.48–11.55 ppb year -1 ), alongside pronounced seasonal variability characterized by autumn–winter peaks. Analysis of ERA5 reanalysis data indicates that this seasonality is closely linked to meteorological conditions, with the lowest wind speeds and shallowest boundary layer heights occurring in autumn and winter, limiting near-surface dispersion and favoring methane accumulation. Persistent and statistically significant hotspots are concentrated in western and southern regions–primarily Mangystau and Atyrau (oil and gas emissions), Turkistan (irrigated agricultural sources), and Kyzylorda (biogenic emissions from dried Aral Sea basins) – as well as localized natural sources in Aktobe. Comparison of the percentile-based and IQR methods indicates that core hotspots are consistently identified across approaches, while their spatial extent and persistence vary with threshold strictness. These findings offer a basis for future zoning of environmental conditions across the region, improvements to air-quality monitoring frameworks, and risk assessment in industrial zones where observational infrastructure remains limited.