Satellite-based technology assessing Ukraine’s ecology under the war

The Donbas is a region dominated by heavy industry, particularly coal mining, chemical processing sites and metallurgy. Intensive mining and steel smelting led to substantial environmental damage before the armed conflict. The fighting in eastern Ukraine has exacerbated an already fragile ecological situation, introducing a range of new risks – of stray munitions hitting extensive chemical and industrial facilities or interconnected mines being flooded and poisoning the water throughout the region. Besides, one of the main hazards is flooding territories associated with closing coal mines. Assessment impacts of uncontrolled mine flooding and towns and villages and associated hazardous processes of surface and groundwater contamination, surface subsidence and dangerous deformations of residential and industrial buildings and other facilities carried out the remote sensing and geological data.


Introduction
Nowadays, the negative anthropogenic impact on the environment is felt especially acutely and has acquired a global character.This is because the effect on nature was previously insignificant, but now it predominates, and new factors constantly appear.Damage caused to natural components leads to negative consequences and reflects the reverse reaction of this impact (adverse for society).Such dangerous environmental changes (primarily of anthropogenic nature) spreads over several hundred thousand square kilometres of the territory of Ukraine.
Recently, the environmental situation in the Donbas has been sharply aggravation caused by the decommissioning of mines and their transition to a new ecological state.We should note that the environmental situation in the Donetsk and Luhansk regions has significantly worsened in recent years and became the theatre of military conflict with active hostilities in 2014-2023 and was partially under the power of the occupation forces, acquiring signs of an ecological disaster.There are 220 coal mines on the territory of the Donetsk coal basin, 97 of which are in operation, 14 are in drainage mode, 39 are in the process of flooding, and 70 are at the liquidation stage.In addition, a total of 2 mines were deregistered due to the completion of liquidation works [1].
Stopping the operation of drainage and ventilation equipment causes the accumulation of water and methane in the mines.The rise in the level of underground water significantly increases the impact of negative artificial factors and leads to the following: flooding of significant areas of pollution of underground and surface sources of drinking water, rivers and reservoirs; a critical reduction in the strength and bearing capacity of mining rocks and soils within the limits of the 1254 (2023) 012115 IOP Publishing doi:10.1088/1755-1315/1254/1/012115 2 influence of mining; expansion of subsidence zones of the massif with the possibility of violations of surface structures; increasing seismic hazard; the formation of unpredicted explosive areas due to changes in methane migration routes.Critical conditions have been created for solving the mentioned problems in almost the entire region of Donbas.
The actual rates of filling mines with water are much higher than expected, which were calculated based on natural filtration coefficients of the mining massif and other multiparametric factors [2].Hence, the need for more accurate data on the environment's basic parameters and unforeseen problems arise after the closure of mines.In addition, the situation is complicated by the need for more reliable data on hydrogeological connections between mines.
The remote sensing technique's efficiency, cost-effectiveness and productivity make it reasonable to use them in solving complex geological hydrogeological, geophysical and environmental problems related to regional geo-ecological security.In this study, ecological issues associated with the flooding of mine waters due to the closure of coal mines are solved (in the example of the territory of the Main Anticline of Donbas).
2. Analysis of the state of the study of the existing native experience regarding the threats of flooding associated with the closure of coal mines Numerous works by Ukrainian scientists give environmental monitoring and coverage of the consequences of the closure of mines.As evidenced by studies of Rudko and Yakovlev [3,4], Dovgiy et al [5,6], Dovgyi and Korzhnev [7], Shybetskyi et al [8], Panova et al [9], Goshovskyi and Zurian [10], Trofymchuk et al [11,12], Anpilova et al [13], Dolin et al [14], Lyuta and Sanina [15].During the full-scale development of the mining area, the geological environment becomes the primary "depot" of the majority of man-made emissions and mechanical, physical and chemical effects on the components of the environment (ground atmosphere and surface hydrosphere, soils, the upper zone of the lithosphere, etc.).
The wet method of closing mines is most often used today in Ukraine -the damp process conserves 90 percents of mines.However, many environmental problems have arisen due to its use since the man-made load on the geological environment; the hydrosphere increases significantly when the mining works are flooded.In addition, large areas of coal-bearing rocks opened by mining and a sharp increase in their permeability due to the man-made fracturing of rocks led to the active influence of mines on the hydrogeological conditions of the surrounding territories [16].
The consequences of mine liquidation have yet to be thoroughly studied, so it isn't easy to choose the final, most rational method of mine liquidation from the point of view of ecology and economic market relations.Nevertheless, the processes of man-made influence must be monitored and analyzed in detail to ensure the minimum negative impact of liquidated mines on the environment.Therefore, it is necessary to carry out routine monitoring observations using ecological methods and the application of remote sensing data.This makes it possible to carry out work in the monitoring mode, to investigate two typical signs of the flooding process -subsidence of the surface and an increase in the area of soil moistening and the release of underground water to the surface.Unfortunately, such studies are still rare in Ukraine.
In particular, modelling the potential development of flooding was carried out, taking into account the zones of propagation of large-amplitude movements, which are given by Ulytsky and Boyko [17].Therefore, the methods of representation of the remote sensing data and the results of the processing are tools for anticipatory forecasting of changes in the hydrogeological situation within the territories of the distribution of mines flooded and liquidated.

Geological structure of the region of the Main Anticline of Donbass
The main anticline is located in the axial part of the Donbas and is the most significant linear fold of the Donetsk Basin.The anticline is confined to the longitudinal Central-Donetsk deep-seated fault, common to Donbas and the Dnipro-Donetsk depression (figures 1 and 2).The main anticline is a linear fold that extends in a straight line along the azimuth of approximately SW 3000 and dips undulipodia to the northwest at an angle of 3-50.Anticlinal folds of the second order are established within its boundaries.The formation of ore bodies, including the Mykytivske mercury deposited, is connected with these structures.
Rocks that make up the fold flanks lie symmetrically with steep dips (50-700).The axial plane of the anticline is vertical or dips steeply to the northeast.Immovables complicate the structure of the Main anticline.
The oldest rocks are exposed in the axial part of the anticline and belong to the Serpukhovian stage of the Lower Carboniferous (geological suite C51), about 400 m thick.The sediments are represented mainly by shale and are characterized by extremely low carbon content.
Deposits of the middle part of the coal system are the most widely distributed in the research region.They are divided into seven suites belonging to the Bashkirian and Moscovian Stages.According to the lithological composition of the rocks, they are typical for the Donetsk basin.This alternation of marine and continental deposits, clay and sandy shales, sandstones and subordinate layers of limestone and coal.The thickness of the Middle Carboniferous deposits increases from 2525 m in the northwest to 4885 m in the southeast.
The middle section is the main coal-bearing part of the Donetsk Carboniferous and has about 66 coal seams, a large amount of which reaches the working thickness.
The lower layers of the Middle Carboniferous rocks consistently cover the sediments of the upper part of the coal system.They are represented by a rhythmic layering of sand-clay rocks,  limestones and coal.However, the frequency and amplitude of oscillations are significantly reduced compared to the underlying layers of rocks, and the section becomes more uniform.Deposits are widely distributed to the southwest and northeast of the axis of the Main anticline within the synclines.
Sediments of the Mesozoic-Cainozoic age are spread outside the Main anticline on the territory of adjacent syncline structures.
The territory where the Main anticline is located belongs to the open Donbas, rising at an average rate of 2.5-3.8mm/year.
According to the data of deep seismic sounding (DSS), a rigid plate of the Precambrian folded basement is reliably fixed under the open Donbas.Thus, the area has a two-tiered structure typical for a platform with imitation of basement faults in the form of tense anticlinal folds and rigid blocks -in the state of fields of relatively calm deposits of Paleozoic rocks with sustained thickness and stable composition.The change in the structure, composition and metamorphism of rocks does not occur gradually or imperceptibly but rather abruptly within areas that are sufficiently narrow and strictly fixed for the entire section.The formed fault-block structure of the territory significantly impacted the development of a wide variety of processes -geodynamic, seismic, filtration, ore and gas formation [19][20][21].
The formation of the Main anticline and the surrounding area, according to Gintov et al [22], took place in 10 stages [23].First, we established that the direction of the tension axis was longitudinal relative to the anticline, and the sub-vertical orientation of the compression axis characterized the Kimmerian stress field.The voltage field has a pulsating character.At the same time, there was a change in the indexing of the tension axis with the intermediate axis of the stress field.The reset character of the deformations in stretching conditions along one axis prevailed [23,24].
Within the Main anticline, the coal-bearing of the southern flank of the anticline is higher than that of the northern flank.Consequently, coking grades of coal have been developed in most of the regions of the district.The degree of coalification in the section increases from the upper to lower horizons.It is represented most clearly on the northern flank and less on the southern side.
The water-bearing capacity of coal deposits is generally low.Inflows into mining operations for individual coal seams are at most 100 m 3 /hour.With depth, the fracturing and waterbearing capacity of coal deposits decreases significantly.At a depth of 400-500 m and more, the deposits of the coal system are practically waterless, except for regions of local irrigation within the zones of tectonic disturbances.
The main aquifers are adapted to limestones and sandstones of the coal system and, within the weathering zones, water-bearing and fractured shales.Limestones are the most fractured rocks.Sources with a flow rate of up to 11.0 l/sec are adapted to their exits to the daylight surface.The flow rates of springs fed by sandstone waters are usually one l/sec.But the total costs for sandstone springs are generally much higher than for limestone springs because sandstones are much thicker than limestone.
Among the many numerous tectonic disturbances in the area, especially discharges, are characterized by increased water capacity.
Thrusts, as a rule, could be better watered.Water inflows from tectonic zones quickly decrease over time, and after a few months, they stop altogether.Coal aquifers are mainly fed by atmospheric precipitation, which causes significant fluctuations in inflows to mines by two or more times throughout the year.In addition, a hydrographic network and numerous mining operations drain coal aquifers.The waters of coal deposits, regardless of water-bearing rocks (limestones or sandstones), are characterized by high mineralization and hardness and therefore are usually unsuitable for drinking and technical water supply.Mine waters have an even worse quality characteristic -the content of dry residue in the water reaches 5000 mg/l and more, and the hardness is more than 100 nem.degree.Despite the poor quality of water from coal deposits, the water is partially used for drinking and technical water supply, and in agriculture, it is used for watering livestock.

Formation of the fractographic base, selection of the necessary satellite images
The database of input materials was formed using the Remote Sensing Software Erdas Imagine, which allowed them to be analyzed and compared.As a coordinate system, a rectangular coordinate system was chosen -UTM / WGS 84 / zone 37.The choice of this coordinate system is due to the zone in which the research area is located (zone 37), and the selection of the UTM / WGS 84 system is because the data of remote sensing of the Earth have precisely this coordinate system.Geometric correction of the input data was carried out in the space image processing program Erdas Imagine using the Geometric Correction module.
The study used a digital terrain model from the Shuttle satellite to the study area.Also, we took satellite images from the Landsat 8 satellite on October 9, 2020, for the study area (figure 3).
Using the Erdas Imagine software, we made a mosaic of topographic maps with a spatial resolution of 17 m.A mosaic of geological maps on a scale of 1: 200,000 was made for the study area (figure 4).

Results of the assessment of the possibility of establishing flooding of territories based on remote sensing data
Mainly we gave attention to the problem of flooding the territory of Donbas to establish the depth of underground water, the places of its exit to the surface, the flow rate and composition of water, the impact on the environment, etc.But we needed more than the available information for a reliable mapping of the areas of flooding of the daylight surface, which was crucial in solving the issues of the ecology of the surrounding territory.All the more so, since we did not carry out the work in the monitoring mode, all the researchers pointed out the need for such work.The interferometry method based on remote sensing data helped to partly solve this issue [17].These materials made it possible to establish the area of flooding of the daytime surface based on the assessment of the subsidence of the Earth's surface, which accompanies the flooding process.Using space images allows for monitoring the development of the flooding process over time.
But the proposed approach does not record the presence of water on the surface or does not allow the estimation of soil moisture.The presence of water or increased soil moisture directly reflects the development of the process of flooding the territory.One effective method of determining the daytime surface's humidity based on remote sensing data is using various water indices.Water indices are designed to assess the presence of moisture in vegetation or open ground.The presence of moisture in vegetation directly indicates its condition: a high moisture level characterizes healthy vegetation characterized by good growth.The humidity of the open ground is directly relevant to our research, while the moisture of the  [19,20]) for the study area, the spatial resolution of 10 m. vegetation cover is an indirect indicator of soil moisture.Therefore, we hope the soil moisture indicator will be more informative than the vegetation cover moisture [25,26].
To calculate humidity indices, values of spectral brightness in the near, medium and infrared ranges are most often used -in the ranges where the radiation is most intensively absorbed by water.These ranges are most sensitive to moisture in vegetation and ground cover.In this study, the five most widely used humidity indices were used: DSWI (Disease water stress index) [27]; NDII (Normalized Difference Infrared Index) [28]; SR-SWIR (Simple Ratio SWIR) [29]; RDI (Ratio Drought Index) [30,31]; NWI (Normalized Water Index) [32].The calculation of these indices was carried out using materials from the Landsat-8 satellite.
To assess the possibilities of using the indicated moisture indices for the study of flooding of the territory of Donbas, a mapping scheme (figure 5) of the potential development of subsurface flooding, which accompanies the flooding of mines, taking into account the propagation zones of large-amplitude movements and numerical modelling of geofiltration, was used [17].
According to the data of the mapping scheme, we selected two areas with depths of underground water levels of 0-3 m (flooded "wet" regions) and more than 10 m from the daytime surface ("dry" regions).The values of the five specified humidity indices were calculated for these areas.Separately, the values of the moisture indices were calculated within the areas of "wet" and "dry" areas covered with woody vegetation and open grounds.The comparison of two distributions of values was carried out using the λ criterion (Kolmogorov-Smirnov) [33].
The maximum values of the λ criterion (table 1) were calculated for dry and flooded areas, and we made the ranking of humidity indices.
The analysis of the obtained results proved that statistically, all the areas that were compared using all five moisture indices differ with reliability higher than 0.99.This allows us to state that we can use multiple satellite imagery to establish flooded areas covered with woody vegetation and open ground.The values of the lambda criterion for multiple and areas with open ground   1).Areas with dry and flooded soils disagree much more strongly than woody vegetation in arid and flooded areas.
The SR-SWIR index will be the most informative among the five studied moisture indices for areas covered with woody vegetation and open ground.The informativeness of other indexes for the territory covered with woody vegetation will be of the same order, much less than the value of the SR-SWIR index.In the region with open ground, all indices are characterized by relative values of informativeness, and only the NWI index is characterized by significantly lower informativeness.

Conclusion
Assessing flooding of the territory, we proposed to use known humidity indices, which are established based on multiple satellite imagery.The analysis of the possibilities of using five moisture indices -DSWI, NDII, SR-SWIR, RDI, and NWI proved that we could use all of them to assess the flooding of territories.Comparing the distributions of the values of the specified indices within dry areas where the depth of groundwater is more profound than 10 m and flooded areas where the depth of groundwater is 0-3 m was performed using the well-known statistical λ criterion (Kolmogorov-Smirnov).As a result of the comparison, it was established that all five water indices could be used to select flooded areas.The ranking of humidification indices according to their capabilities, which was based on the analysis of the values of the λ criterion, showed that the SR-SWIR index is characterized by the most significant capabilities, both for areas covered with woody vegetation and for areas with open ground.
Therefore, the possibility of using multi-zone space images to assess the flooding of the territory of Donbas due to the mass closure of coal mines through the so-called "wet conservation" has been established.

Table 1 .
Maximum values of the λ criterion on dry and flooded areas covered with woody vegetation and areas with open ground at different moisture indices.