Macro-scale earthquake forecasting on spatially discretized grids presents a severe class imbalance problem, where rare seismic events are embedded within long periods of inactivity. In Central Asia, forecasting \((M \ge 3.0)\) earthquakes on a weekly \((1^\circ \times 1^\circ)\) grid produces approximately 96% zero inflation and an event prevalence of about 4.5%, yielding a constant Precision--Recall Area Under the Curve (PR-AUC) baseline of 0.045. This study investigates whether predictive performance under such extreme imbalance is governed primarily by model architecture or by structured feature design and baseline calibration. We introduce a framework that integrates tectonic regime--conditioned normalization, Omori energy decay proxies, structural fault descriptors, and Log-Odds baseline initialization that encodes historical cell-specific event rates directly into the learning objective. Across six heterogeneous architectures evaluated under a strict chronological split, performance converges to PR-AUC \((\approx 0.451)\) and Receiver Operating Characteristic Area Under the Curve (ROC-AUC) \((\approx 0.844)\), representing a tenfold improvement over the constant baseline without synthetic resampling. The results indicate that under extreme low-prevalence conditions, calibrated baselines and physically structured feature spaces dominate architectural complexity, suggesting the existence of a macro-scale predictability ceiling in seismic forecasting.

Rock glaciers are key indicators of mountain permafrost and act as climatically resilient water reservoirs in arid mountains. This study presents the first inventory and kinematic classification of rock glaciers in Western Tien Shan (Kazakhstan and Kyrgyzstan), combining geomorphological mapping with InSAR time-series analysis. Using high-resolution optical imagery (Google Earth Pro (version 7.3.6.10441), Bing Maps, SAS Planet (version 200606.10075), digital elevation models, and Small Baseline Subset InSAR processing, 741 rock glaciers covering more than 70.5 km(2) were identified. Activity classification revealed 232 transitional and 509 active forms, with mean seasonal displacement rates of similar to 15 cm yr(-1) calculated based on August and September observations. Spatial analysis showed a strong rock glacier concentration on north-facing slopes (>66% of total area) with reduced potential incoming solar radiation. Rock glaciers mainly occur between 2800 and 3800 m a.s.l., with a mean elevation of 3340 m a.s.l. However, their kinematic activity varies across mid-altitudinal ranges, underscoring the influence of slope, aspect, shading, and local topography. Integration with the Global Permafrost Zonation Index (PZI) indicated a lower permafrost boundary at similar to 1922 m a.s.l., with the largest and most active glaciers occurring at intermediate PZI values (0.5-0.7). This first rock glacier inventory for the Western Tien Shan establishes a benchmark dataset that supports the validation and refinement of global models at a regional scale, guides priorities for permafrost monitoring, and provides a replicable framework for inventory development in other data-scarce mountain regions.

Monitoring the stability of hydraulic structures such as dams and reservoirs in seismically active regions is essential for ensuring their safety and operational reliability. This study presents a comprehensive geospatial approach combining lineament analysis and geodynamic zoning to assess the structural stability of the Voroshilov and Priyut reservoirs located in the Almaty region, Kazakhstan. A regional lineament map was generated using ASTER GDEM data, while ALOS PALSAR data were used for detailed local analysis. Lineaments were extracted and analyzed through automated processing in PCI Geomatica. Lineament density maps and azimuthal rose diagrams were constructed to identify zones of tectonic weakness and assess regional structural patterns. Integration of lineament density, GPS velocity fields, InSAR deformation data, and probabilistic seismic hazard maps enabled the development of a detailed geodynamic zoning model. Results show that the studied sites are located within zones of low local geodynamic activity, with lineament densities of 0.8–1.2 km/km2, significantly lower than regional averages of 3–4 km/km2. GPS velocities in the area do not exceed 4 mm/year, and InSAR analysis indicates minimal surface deformation (<5 mm/year). Despite this apparent local stability, the 2024 Voroshilov Dam failure highlights the cumulative effect of regional seismic stresses (PGA up to 0.9 g) and localized filtration along fracture zones as critical risk factors. The proposed geodynamic zoning correctly identified the site as structurally stable under normal conditions but indicates that even low-activity zones are vulnerable under cumulative seismic loading. This demonstrates that an integrated approach combining remote sensing, geodetic, and seismic data can provide quantitative assessments for dam safety, predict potential high-risk zones, and support preventive monitoring in tectonically active regions.

Glaciers in the Northern Tien Shan are a major source of Ile River runoff, supplying water to Kazakhstan's largest city, Almaty. Under ongoing climate warming, their degradation alters the magnitude and seasonality of river discharge, increasing water-resource vulnerability. This study quantifies long-term changes in glacier area, firn-line elevation, and glacier runoff in the northern Tien Shan from 1955 to 2021. The analysis uses multi-decadal meteorological observations, hydrological records, multi-temporal Landsat-7/8 and Sentinel-2 imagery, and DEMs combined with empirical and semi-empirical runoff estimation methods. The glacier area has declined by more than 45-60% since 1955, accompanied by a rise in firn-line altitude from similar to 3400 to 3700 m. At the Mynzhylky station, mean summer air temperature increased by 0.88 degrees C, reflecting persistent warming in glacierized elevations. The contribution of glacier meltwater to annual discharge decreased from similar to 32% in 1955-1990 to similar to 25% in 1991-2021, while total and vegetation-period runoff increased due to modified seasonal hydrological conditions. These results demonstrate the impact of climate warming on glacier-fed runoff in the Northern Tien Shan and highlight the need to integrate glacier degradation into water-resource management and long-term water-security assessments.

Glacial-lake outburst floods (GLOFs) threaten more than three million residents of south-east Kazakhstan, yet quantitative data on lake growth and storage are scarce. We inventoried 154 lakes on the northern flank of the Ile-Alatau and selected four moraine-dammed basins with the greatest historical flood activity for detailed study. Annual lake outlines (2016–2023) were extracted from 3 m PlanetScope imagery with a Normalised Difference Water Index workflow, while late-ablation echo-sounder surveys (2023–2024) yielded sub-metre bathymetric grids. A regionally calibrated area–volume power law translated each shoreline to water storage, and field volumes served as an independent accuracy check. The lakes display divergent trajectories. Rapid thermokarst development led to a 37% increase in the surface area of Lake 13bis, expanding from 0.039 km2 to 0.054 km2 over a 5-year period. In contrast, engineering-induced drawdown resulted in a 44% reduction in the area of Lake 6, from 0.019 km2 to 0.011 km2. Lakes 5 and 2, which are supplied by actively retreating glaciers, exhibited surface area increases of 4.8% and 15%, expanding from 0.077 km2 to 0.088 km2 and from 0.061 km2 to 0.070 km2, respectively. The empirical model reproduces field volumes to within ±25% for four lakes, confirming its utility for rapid hazard screening, but overestimates storage in low-relief basins and underestimates artificially drained lakes. This is the first study in Ile-Alatau to fuse daily 3 m multispectral imagery with ground-truth bathymetry, delivering an 8-year, volume-resolved record of lake evolution. The results identify Lake 5 and Lake 2 as priority targets for early-warning systems and demonstrate that sustained intervention can effectively suppress GLOF risk. Incorporating these storage trajectories into regional disaster plans will sharpen evacuation mapping, optimise resource allocation, and inform transboundary water-hazard policy under accelerating climate change.

High-mountain regions such as the Ile-Alatau range of the Northern Tien Shan are highly sensitive to climatic fluctuations, where even minor variations in temperature and precipitation significantly influence glacier mass balance and hydrology. Despite this sensitivity, few long-term studies have examined the links between hydro-climatic trends, glacier retreat, and moraine lake development. This study investigates multi-decadal glacier and lake dynamics (1955–2024) in relation to observed climate variability, using an integrated hydro-climatic and remote-sensing approach. Temperature and precipitation records from four high-altitude meteorological stations were assessed using linear regression and the Mann–Kendall test, while glacier and lake extents were derived from aerial photographs and Landsat, Sentinel-2, and PlanetScope imagery across ten river basins. Results show statistically significant warming at all stations, with mean annual temperatures increasing by 0.14–0.28 °C per decade and summer temperatures by 0.15–0.30 °C, while precipitation remained stable or slightly decreased. Glacierized area decreased from approximately 269.6 km2 in 1955 to 141.7 km2 in 2021, representing a 47.4% reduction (≈−0.72% yr−1) over six decades and underscoring the rapid regional cryospheric response to sustained climatic warming. Simultaneously, moraine-dammed lakes increased by 16–18% between 2017 and 2024. These trends highlight the dominant climatic control on glacier loss and lake evolution, emphasizing growing glacial lake outburst floods (GLOFs) and the need for adaptive water-resource management in Central Asia.

Assessing glaciers using recent and historical data and predicting the future impacts on them due to climate change are crucial for understanding global glacier mass balance, regional water resources, and downstream hydrology. Computational methods are crucial for analyzing current conditions and forecasting glacier changes using remote sensing and other data sources. Due to the complexity and large data volumes, there is a strong demand for accelerated computing. AI-based approaches are increasingly being adopted for their efficiency and accuracy in these tasks. Thus, in the current state-of-the-art review work, available research results on the application of AI methods for glacier studies are addressed. Using selected search terms, AI-based publications are collected from research databases. They are further classified in terms of their geographical locations and glacier-related research purposes. It was found that the majority of AI-based glacier studies focused on inventorying and mapping glaciers worldwide. AI techniques like U-Net, Random forest, CNN, and DeepLab are mostly utilized in glacier mapping, demonstrating their adaptability and scalability. Other AI-based glacier studies such as glacier evolution, snow/ice differentiation, and ice dynamic modeling are reviewed and classified, Overall, AI methods are predominantly based on supervised learning and deep learning approaches, and these methods have been used almost evenly in glacier publications over the years since the beginning of this research area. Thus, the integration of AI in glacier research is advancing, promising to enhance our comprehension of glaciers amid climate change and aiding environmental conservation and resource management.

While rock glaciers (RGs) are widespread in the Zhetysu Alatau mountain range of Tien Shan (Kazakhstan), they have not yet been systematically investigated. In this study, we present the first rock glacier inventory of this region containing 256 rock glaciers with quantitative information about their locations, geomorphic parameters, and downslope velocities, as established using a method that combines SAR interferometry and optical images from Google Earth. Our inventory shows that most of the RGs are talus-derived (61%). The maximum downslope velocity of the active rock glaciers (ARGs) was 252 mm yr−1. The average lower height of rock glaciers in this part of the Zhetysu Alatau was 3036 m above sea level (ASL). The largest area of rock glaciers was located between 2800 and 3400 m ASL and covered almost 86% of the total area. Most rock glaciers had a northern (northern, northeastern, and northwestern) orientation, which indicated the important role of solar insolation in their formation and preservation.

An updated glacier inventory is important for understanding the current glacier dynamics in the conditions of actual accelerating glacier retreat observed around the world. Here, we present a detailed analysis of the glaciation areas of the Zhetysu Alatau Range (Tien Shan) for 1956–2016 using well-established semiautomatic methods based on the band ratios. The total glacier area decreased by 49 ± 2.8% or by 399 ± 11.2 km2 from 813.6 ± 22.8 km2 to 414.6 ± 11.6 km2 during 1956–2016, while the number of glaciers increased from 985 to 813. Similar rates of area change characterized the periods 1956–2001, 2001–2012, 2012–2016, and 2001–2016: −296.2 ± 8.3 (−0.8% a−1), −63.7± 1.8 (−1.1% a−1), −39.1 ±1.1 (−2.2% a−1) and −102.8 ± 2.9 (−1.3% a−1) km2, respectively. The mean glacier size decreased from 0.57 km2 in 2001 to 0.51 km2 in 2016. Most glaciation areas of the Zhetysu Alatau faced north (north, northwest, and northeast), covered 390.35 ± 11 km2, and were located in altitudes between 3000 and 4000 m.a.s.l. With shrinkage rates of about −0.8% and −1.3% a−1 for the periods of 1956–2001 and 2001–2016, our results show that study area has the highest shrinkage rate compared to other glacierized areas of Central Asian mountains, including Altai, Pamir, and even the inner ranges of Tien Shan. It was found that a significant increase in temperature (0.12 °C/10 years) plays a main role in the state of glaciers.

This article discusses the space monitoring system, its composition and functional diagram of the basic elements of the system and remote access services to the resources and results of the space monitoring system. The composition of the information support of the system, its formats and sources, and the frequency of updates are described. The basic structure of the geoportal, namely, the subsystem for collecting satellite and ground data, a database, processing and expert interpretation, integration and dissemination of monitoring data, is briefly described. The whole cycle of space monitoring work is described for each priority area, from obtaining raw data to uploading emergency monitoring data to the geoportal in Kazakhstan, including operational data from space monitoring of forest and steppe fires, GPS monitoring of intense movements of the earth's crust, space monitoring of floods, space monitoring oil pollution of the Caspian Sea with the corresponding "screenshots". The geoportal interface is described.