Since February 2022, the full-scale war in Ukraine has been strongly affecting society and economy in Ukraine and beyond. Satellite observations are crucial tools to objectively monitor and assess the impacts of the war. We combine satellite-based tropospheric nitrogen dioxide (NO2) and carbon dioxide (CO2) observations to detect and characterize changes in human activities, as both are linked to fossil fuel combustion processes. We show significantly reduced NO2 levels over the major Ukrainian cities, power plants and industrial areas: the NO2 concentrations in the second quarter of 2022 were 15–46% lower than the same quarter during the reference period 2018–2021, which is well below the typical year-to-year variability (5–15%). In the Ukrainian capital Kyiv, the NO2 tropospheric column monthly average in April 2022 was almost 60% smaller than 2019 and 2021, and about 40% smaller than 2020 (the period mostly affected by the COVID-19 restrictions). Such a decrease is consistent with the essential reduction in population and corresponding emissions from the transport and commercial/residential sectors over the major Ukrainian cities. The NO2 reductions observed in the industrial regions of eastern Ukraine reflect the decline in the Ukrainian industrial production during the war (40–50% lower than in 2021), especially from the metallurgic and chemical industry, which also led to a decrease in power demand and corresponding electricity production by thermal power plants (which was 35% lower in 2022 compared to 2021). Satellite observations of land properties and thermal anomalies indicate an anomalous distribution of fire detections along the front line, which are attributable to shelling or other intentional fires, rather than the typical homogeneously distributed fires related to crop harvesting. The results provide timely insights into the impacts of the ongoing war on the Ukrainian society and illustrate how the synergic use of satellite observations from multiple platforms can be useful in monitoring significant societal changes. Satellite-based observations can mitigate the lack of monitoring capability during war and conflicts and enable the fast assessment of sudden changes in air pollutants and other relevant parameters.
Since the beginning of the full-scale war in Ukraine in February 2022, the life of the Ukrainian population and the country’s economy have been strongly affected by the war. By June 2022 Russia occupied 20% of Ukraine's territory1. Almost 19,000 civilian casualties have been registered in the country after 1 year of war2. The United Nations High Commissioner for Refugees (UNHCR) reported that about 8 million Ukrainians (about 24% of the total population) left the country as refugees across Europe3. The World Bank estimates the reconstruction and recovery cost as $349 billion, which is more than 150% of Ukraine’s GDP in 20214. War-induced environmental damages and risks to energy and industrial infrastructure, urban areas, agricultural and natural resources have also been discussed5,6,7. Reductions in crop harvesting have been reported near the front line8, which impacts global food security, as Ukraine is a major world exporter of wheat9. Despite the ongoing efforts by the Ukrainian authorities10 and other organizations11,12, the monitoring capabilities have been significantly limited due to lack of data and incomplete information, especially from the occupied areas, and the uncertainties associated with the impact assessments remain large. Also, the emissions from military actions are difficult to assess and are not included in traditional emission inventories13.
Satellite-based Earth observations can collect near real-time objective information globally including areas where ground-based data collection is difficult. Satellite observations of air pollutants such as nitrogen dioxide (NO2) have a history of success in monitoring the air quality levels and their improvements globally14. Nitrogen oxides (NOx = NO + NO2) are mostly emitted from fossil fuel combustion processes, e.g., from coal-burning power plants, metallurgic and coke production processes, transportation, and fires. Satellite NO2 observations have been used to derive NOx emissions and their trends, and to evaluate the success of environmental policy measures and the impacts of the global economic crises15,16,17,18,19. Such observations have also been applied to rapidly assess NO2 pollution levels over several cities in response to the reduced mobility and industrial activities during the COVID-19 pandemic20,21,22,23. Reductions in NO2 and ammonia have been previously reported as a consequence of the Syrian Civil War in 201124,25. Similarly, over the past decade, satellite observations of greenhouse gasses (GHGs), such as CO2 and CH4, have become available and started playing a crucial role in climate mitigation monitoring (26 and reference therein). Recent studies have demonstrated that the combined use of satellite air quality and GHG data improves the ability of estimating surface emissions and characterizing the sources27,28,29. Combining a wide variety of satellite observations, including atmospheric concentrations and imagery data can further enhance the ability to monitor changes and attribute them to specific human and/or natural processes.
In this work, we exploit the synergy of different Earth satellite observations from multiple platforms to study the impacts of the war in Ukraine. We use the Copernicus Sentinel 5-Precursor/TROPOspheric Monitoring Instrument (S5P/TROPOMI) NO2 tropospheric column data30 to assess the variations in fossil fuel usage related to changes in population, power generation and other industrial activities in Ukraine. We also explore satellite imageries as well as fire/thermal anomaly data to characterize the signatures and the spatio-temporal distribution of combustions (e.g., shelling and other fires) near the front line. We also consider column-averaged CO2 dry air mole fraction (XCO2) data, collected from NASA’s OCO-2 (Orbiting Carbon Observatory-2) satellite to analyze persistent CO2 anomalies.
Substantial NO2 reductions observed over Ukraine
We observed substantial NO2 reductions over many major urban and industrial areas in Ukraine (Supplementary Fig. 1) from March to August 2022 as compared to previous years (Fig. 1). The observed reductions of NO2 are consistent in timing and location with the reported reductions in traffic, commercial and residential activities in the cities as well as from slowed down power and heat production and industrial activities during the war31. Correspondingly, muted nighttime lights (NTL) have been observed over several Ukrainian cities32,33 by the VIIRS satellite nighttime imagery from the Day/Night Band (DNB, 500–900 nm). NO2 reductions as compared to previous years are observed for example over the capital city of Kyiv (Supplementary Fig. 2) and in the area around the city of Donetsk in south-eastern Ukraine (Supplementary Fig. 3). Kyiv was attacked during the first phase of the war (February–March 2022) and a significant part of the population left the city at that time but partially returned by July 202234. Near the city of Donetsk, the largest NO2 enhancements correspond to five coal-fired power plants (Supplementary Fig. 3). This area was partially occupied by Russia in 2014 (Supplementary Fig. 1). Two power plants are controlled by Russia (Starobeshivska and Zuyivska, with capacity of 2010 MW and 1220 MW, respectively) and two by Ukraine (Kurakhivska and Slovyanska, with capacity of 1532 MW and 830 MW, respectively). In addition, the Vuhlehirska power plant (capacity 3600 MW) was occupied by Russia on 26 July 202235. The NO2 levels observed in 2022 over the power plants controlled by Russia since 2014 remain closer to those of the previous years, while the other facilities experienced more substantial reductions. These large coal-fired power plants are located a short distance from the front line and might have experienced disruptions already since 2014, but possible changes in NO2 before 2022 remain closer to the normal year-to-year variability. In 2021, there was also a period with critically low coal reserves at the thermal power plants controlled by Ukraine, which did cause a decrease in electricity generation for example from the Vuhlehirska power plant36. There are several large metallurgical and coke plants near Donetsk (see Supplementary Fig. 1) but they had a lower impact on the overall NO2 patterns (Supplementary Fig. 3).
The NO2 levels were also strongly reduced over the Dnipro-Zaporizhzhia-Kryvyi Rih region (Supplementary Fig. 4), including the contribution from Zaporizhzhia’s industrial area (in particular Zaporizhkoks coke and ArcelorMittal steel plant); Enerhodar, where the coal-burning Zaporizka thermal power plant is located; Kryvyi Rih, including coke and the Inhuletskyi ore beneficiation plant in the south and the Ternivskyi District in the north (the largest in Europe Pivnichnyi ore beneficiation plant, and Thentralnyi ore beneficiation plant); Dnipro city (steel and cast iron production); Kamianske, hosting both heavy and chemical industry; Horishni Plavni, a purpose-built iron mining city. The NO2 difference map (Fig. 2a) further highlights the localized reductions in NO2 levels in 2022 as compared to 2021 over the impacted areas.
In 2022 the quarterly mean NO2 concentrations were well below the quarterly mean levels from the reference period (2018–2021) and the NO2 reductions were larger than the typical variability (5–15%) at most locations (Fig. 3). The NO2 concentrations over the major cities and emission areas in the second quarter of 2022 were 15–46% lower than the same quarter during the reference period 2018–2021 (Table 1). A slight gradual reduction in the NO2 levels is apparent since 2018 at some locations (Fig. 3). In addition to changes in emissions, the observed year-to-year variability can partly depend on meteorological factors, especially related to changes in wind conditions21. When considering only calm wind conditions (e.g., wind speed smaller than 3 m/s), which reduces the effect of NO2 outflow away from the source, we find that the NO2 reductions in the second quarter of 2022 become even larger (Supplementary Fig. 5, Table 1) as compared to the reference period. One exception is the city of Lviv, for which a 3% increase in NO2 concentrations is observed in the second quarter of 2022 under calm wind conditions (instead of 15% reduction under all wind conditions). Lviv, located near the western border with Poland, did not experience reductions in population as other Ukrainian cities and was hosting 240,000 internally displaced persons as of July 202237. Further variability is possibly related to data availability (Supplementary Fig. 6), which is mostly driven by differences in cloud coverage.
The analysis of NO2 observations available from the Ozone Monitoring Instrument (OMI) enabled the comparison to a longer reference period before the COVID-19 pandemic (2015–2019). Based on OMI observations, Supplementary Fig. 7 (solid lines) shows that the 2022 spring–summer mean NO2 concentrations over the most polluted areas in Ukraine were well below the 2015–2019 period (− 32% in Kyiv, − 23% near Donetsk and − 24% in the Dnipro/Zaporizhzhia area) and below the normal year-to-year variability (10–12%). Also, considering the short reference period 2019–2021 (dashed lines in Supplementary Fig. 7) for OMI observations, only changes the observed reduction in 2022 by 1 percentage point.
The observed NO2 reductions persisted for several months in 2022, unlike the short-term reduction observed during the COVID-19 restrictions in April 2020 (Fig. 4). The NO2 tropospheric column monthly averages over Kyiv in April 2022 were still about 40% smaller than those in 2020, and almost 60% smaller than 2019 and 2021. We note that changing meteorological conditions and data availability can affect the 15-days running mean NO2 values more than the quarterly NO2 anomalies. Also, the NO2 data availability is poorer during winter months due to lack of sunlight and frequent cloudy conditions, which yields higher variance in the running means (Fig. 4). The NO2 reductions reported here are generally larger than previously reported values38, possibly because we use consistent algorithm versions of the TROPOMI NO2 product throughout the study period.
Overall, we observed NO2 reduction over several cities, power plants and industrial areas in Ukraine. Some (Enerhodar, Mariupol, Vuhlehirska power plant) were occupied, others were close to the front line (Supplementary Fig. 1). The decrease in human activity during the war reduced electricity demand, which in turn reduced power plants’ production and thereby reduced NOx emissions. In the bigger cities, the decrease in population has been likely driving the reduction in emissions from the transport and residential/commercial sector. A significant part of the population left the cities in the early days of the war, mostly from Kyiv and Kharkiv, and those who remained were subjected to curfew39. Also, freight transportation (as compared to the same period in 2021) decreased by 22% from road transport and 50% from railways40.
The World Bank reported a first estimate of significant economic decline in the country4. Ukraine’s Gross Domestic Product (GDP) declined in 2022 by over 30% year-on-year41. The State Statistics Service of Ukraine reported about 40–50% decline in industrial production in 202242, especially from the steel and chemical industry that is concentrated in eastern Ukraine43. Such decline in industrial production led to a decrease in the use of electricity. According to the Ministry of Energy of Ukraine, Ukrainian industry reduced electricity consumption by 45% in 202244. In particular, the metallurgical industry reduced electricity consumption by 52% (after the loss of the metallurgical giant in Mariupol), the chemical and petrochemical industry reduced consumption by 60%, the production of building materials by 47%, engineering by 38%, the food and processing industry by 24%, fuel industry by 24%, households by 16%. The total production of electricity in Ukraine in 2022 decreased by 27.5% compared to 2021 (production by thermal power plants decreased by 35%, and combined power-heat plants by 32%). The International Energy Agency also reported reductions in Ukraine’s power generation, including from coal-based energy sources45. These changes are consistent with the observed NO2 reductions, which can be linked to reductions in fossil fuel usage.
Exceptional fire patterns near the front line
We also observe unique fire and thermal anomaly patterns in 2022. We analyzed the Suomi-NPP VIIRS data to further assess which areas are mostly disrupted by the war (Fig. 5). During the period 2019–2021 (before the full-scale war), the detected fires, which are generally attributed to crop harvesting or other vegetation fires, were distributed homogeneously over the Ukrainian territory (Fig. 5a-c) with most fires detected in March–April in north-western Ukraine and in July–August in the south-eastern part of the country. Some of the year-to-year variability is possibly related to changing cloud coverage and data availability. However, in 2022 most of the detected fires were located along the front line in the northern, eastern and southern parts of Ukraine, suggesting possible contribution from shelling (Fig. 5d). The 2022 spatio-temporal patterns are in line with the reported changes in the front line (Supplementary Fig. 1,46). For example, several fires were observed north-west of Kyiv (Supplementary Fig. 8) in the direction of Irpin-Ivankiv during March, when the Russian offensive focused on the area surrounding the Ukrainian capital (Supplementary Fig. 1). The Sentinel 2 false color images confirm the occurrence of several fires around Irpin on 21 March 2022 (Supplementary Fig. 9), just before the city was retaken by the Ukrainian troops 47. Towards the summer the fires are mostly detected along the front line in the eastern and southern parts of the country, which is in line with the location of the frontline at that time (Supplementary Fig. 1).
Satellite imagery data also provide snapshots of drastic changes in the fire or thermal anomaly patterns over the city of Mariupol, which was attacked during the first three months of the war (Fig. 6). In 2021, the Sentinel 2 false color images show thermal anomalies due to the hot smoke from the metallurgic industrial facilities (Fig. 6a), such as Azovstal and Mariupolsky Metal plant, and the Markohim coke plant. However, the thermal anomalies were no longer detected after the beginning of the full-scale war in March 2022 (Fig. 6b). This would suggest that an interruption of industrial activities occurred during the war. This is in line with the NO2 reduction observed in 2022 over the metallurgic industrial complex in the northern part of the city (Supplementary Fig. 10). The observed NO2 reduction can also be partially attributed to the reduced human activities due to the extensive evacuation of the city up until May 2022. On 24 March 2022, when the Russian offensive focused on Mariupol, larger and more intense clusters of fires were observed all around the city and countryside (Fig. 6b).
Similarly, a large cluster of fires was also observed in 2022 over the rural area around Velyka Oleksandrivka, near the front line and temporarily occupied territories north of Kherson (Supplementary Fig. 11b). This area was among those left unharvested due to the persistent shelling near the front line8. The unharvested wheat fields appear in the Sentinel 2 natural color images dark brown on 7 August 2022 (at the end of the harvesting season) as compared to the harvested (light brown) fields to the north and south (Supplementary Fig. 12b). In comparison, during the same day in 2020 the whole area appears homogeneous, without darker spots (Supplementary Fig. 12a), with fewer fires detected between March and August 2020 (Supplementary Fig. 11a), typically related to vegetation fires.
CO2 anomalies indicate reductions in fossil fuel combustion
CO2 observations further explain the observed NO2 reductions over the areas most affected by the war. We find visual correspondence between the spatial patterns of the NO2 and CO2 anomaly differences, especially over areas with significant NO2 reductions. In particular, we observe a East–West gradient in the CO2 anomaly difference maps, with lower CO2 anomalies (as compared to 2021) over the eastern part of Ukraine (Fig. 2b), which is in line with the observed NO2 reductions. Since NOx and CO2 are often co-emitted (e.g., from fossil fuel combustion), paired NO2 and CO2 anomaly reductions as those observed in eastern Ukraine indicate reductions in fossil fuel use in cities and industrial areas. On the other hand, slightly positive or less substantial NO2 and CO2 anomaly changes can be observed over western Ukraine. These differences are less localized than those observed in eastern Ukraine, where the largest emission sources are located, and can have large contribution from natural variability.
Because of the long lifetime, CO2 anomalies are more affected by atmospheric transport and not as localized as NO2. Due to the narrow swath of the OCO-2 instrument (width smaller than 10 km), the observed CO2 anomalies might not be able to fully capture year-to-year changes in CO2 emissions. Nevertheless, the expected reduction in fossil fuel usage is persistent in time over a relatively large region, which makes it possible to sufficiently cover the emission areas.
Additional NOx and CO2 emissions compared to previous years are expected in the areas impacted by fires (Fig. 5d) as the result of military activity such as shelling, fires of petroleum products, buildings, and infrastructure objects, as well as forest fires and fires of agricultural fields. While the actual composition is unknown, military-related combustions could be fuel inefficient and might cause NOx emissions in excess. Relatively small increases in NO2 can be observed in 2022 over some areas north-west of Kyiv, south-east of Kharkiv and in the south-eastern part of Ukraine (Fig. 2a) which were strongly affected by such fires, but the enhancements are difficult to attribute to a specific cause. Also, an increase in CO2 emissions is expected from the fires related to military activities near the front line. Military emissions though might be smaller than the emissions from different economic sectors in previous years, and appear to be partially hidden by the substantial reductions related to fossil fuel combustion in the average maps (Fig. 2) due to their temporary nature and sparse distribution, as well as the sparse coverage of the satellite observations.
Our results provide early insights into the magnitude of the impacts on human and industrial activities in Ukraine during the first stages of the full-scale war, but they remain in part qualitative. In addition to changes in emissions, concentrations of atmospheric gases depend on changing meteorological conditions, and further analysis will be needed to accurately estimate changes in air pollution and GHG emissions due to the war. Nevertheless, satellite-based observations make it possible to monitor such abrupt changes almost in real time, as compared to other methods that require longer time to collect all relevant information.
During peaceful times, NO2 reductions as those reported here would be considered as a welcome improvement of air quality and human health. In this case the observed changes tell a different story about the extent of the disruption caused by the war on the Ukrainian society. The NO2 reductions observed during March-August 2022 over power plants and other industrial areas reflect the reduction in industrial production and electricity demand from several industrial sectors, especially from the metallurgical and chemical industry. Drastic reductions in population (as well as changes in their mobility habits) have been reported for many Ukrainian cities, after many civilians moved to safer areas in the western part of the country or abroad48. The UN projects further population decline in 2023 and beyond49. The reductions in fossil fuel consumption in Ukraine might have been partly offset by an increase elsewhere12.
Overall, our analysis shows how the synergic use of Earth observations from different satellite platforms provides an important asset to monitor significant societal changes and has potential to support authorities in evaluating the impacts of the war on the Ukrainian society based on independent information. As presented in this study, the lack of monitoring capability in such situations as war and conflicts can be mitigated by combining satellite observations of trace gasses, which are now available with unprecedented resolution and accuracy, together with other satellite imagery and products. For example, satellite NO2 data can supplement and enhance NTL as indicator of human activity changes50,51, especially those not associated with lights or that happened during a power outage. While assessing the impacts of socio-economic changes using satellite observations generally requires expert analysis, new platforms and tools for data distribution and analysis are constantly developing and might enable faster and easier assessments in the future, also for non-expert users. Future instruments such as the European Copernicus CO2M (Anthropogenic Carbon Dioxide Monitoring52) and Japan’s GOSAT-GW53 missions will provide collocated observations of NO2 and CO2 which should enhance our ability of assessing changes in human activities during major socio-economic disruptions, as those presented here. Furthermore, atmospheric geostationary missions such as Sentinel-454, Tropospheric Emissions: Monitoring of Pollution (TEMPO55), and Geostationary Environment Monitoring Spectrometer (GEMS56) will provide useful NO2 observations with multiple overpasses during one day.
Finally, the results shown here cover the time period between March-August 2022. Instead, beginning in the autumn of 2022, Russia launched regular missile attacks across the whole Ukraine to damage the energy infrastructure. This, in turn, affects industry and all spheres of life in Ukrainian society. Further studies should focus on the assessment of the full economic and environmental impacts throughout the war as well as during the reconstruction phase.
Materials and methods
TROPOMI and OMI NO2 product description and data processing
Nitrogen dioxide (NO2) is used as a proxy for human activities, as it is linked to fossil fuel combustion processes, and it is often co-emitted with carbon dioxide (CO2). Since NO2 is short-lived in the lower atmosphere, it can be found in high concentrations near the emission sources and can therefore be used to assess changes in emissions. Nitrogen oxides (NOx) are largely emitted as nitric oxides (NO), which rapidly transforms via photochemical reactions into NO2 (the species retrieved by satellite) until reaching a steady state.
TROPOMI is a passive-sensing hyperspectral nadir-viewing imager aboard the Copernicus Sentinel-5 Precursor (S5P) satellite, launched on 13 October 2017. S5P/TROPOMI NO2 retrievals are available at spatial resolution of 3.5 × 5.5 km2 (nadir) and daily global coverage, with overpass local time of 13:30. The NO2 retrievals are derived using TROPOMI’s UV–VIS spectrometer backscattered solar radiation measurements in the 405–465 nm wavelength range. Here, we use the reprocessed S5P-PAL NO2 dataset57, which includes a reprocessing of the TROPOMI NO2 data product from the start of the S5P mission until November 2021 using version 2.3.1 of the retrieval algorithm. Additionally, offline (OFFL) NO2 products are used to cover the period after November 2021.
The TROPOMI NO2 tropospheric columns were used to compile March-August average NO2 maps for the analysis of persistent NO2 enhancements and their year-to-year changes covering the period 2019–2022. TROPOMI NO2 retrievals are gridded into a regular 1 × 1 km2 grid using the area-weighted gridding method. Supplementary Fig. 6 shows the amount of data for each grid cell estimated from the sum of the weights used in the gridding. The weights (in the range 0–1) are defined for each orbit as the fraction of grid cell area covered by valid NO2 observations, with 1 corresponding to the grid cell fully covered. The amount of data is generally larger over water and lower near the mountains. The average maps for 2022 are based on about 21% (13%) more valid observations than for 2019 (2021) and about the same amount for 2020, mostly due to varying cloud conditions. In general, there are less valid NO2 retrievals during wintertime, due to frequent cloudy conditions and lack of sunlight.
We also derive the time series over the major cities and industrial sites in Ukraine, identified by the largest NO2 enhancements in the average maps in Fig. 1. We consider all the good quality TROPOMI NO2 observations within 4 km distance from the site of interest. We then derive quarterly averages anomalies as the difference between the quarterly averages for each year and the corresponding quarterly average for the reference period 2018–2021 (note that only third and fourth quarters are available for the year 2018). The quarterly anomalies are calculated to adjust for NO2 seasonal variability, as NO2 concentrations are generally higher in winter and smaller in summer, due to lower emissions and increased photo-chemistry in the summertime. The distance of 4 km from the source was chosen to avoid signals from neighboring sources to mix. The results do not substantially change when considering larger distance; for example, doubling the distance to the site to 8 km changes the quarterly differences reported in Table 1 by 0–5 percentage points. In order to reduce the effect of changing wind conditions, we calculate the NO2 anomalies considering only TROPOMI observations with corresponding wind speed smaller than 3 m/s (Supplementary Fig. 5). The 10-m wind components from the European Centre for Medium-Range Weather Forecasts—Integrated Forecasting System (ECMWF-IFS) are available within the TROPOMI NO2 product and are sampled according to the satellite pixels. Under calm wind conditions the NO2 enhancements are generally more pronounced and the spatial patterns are closer to the emission patterns, as pollution is not transported far away from the sources. Since satellite NO2 retrievals are only available under clear sky conditions, cloudiness will have a small effect except for changing the data availability (Supplementary Fig. 6).
In order to compare the NO2 levels in 2022 to a longer reference period, we use NO2 tropospheric columns from the Ozone Monitoring Instrument (OMI58) on board NASA’s Aura satellite, to derive spring–summer anomalies during the period 2015–2022 (as compared to the reference periods 2015–2019 and 2019–2021) over three areas of interest: Kyiv, Donetsk and the area around Zaporizhzhia. We use the OMI Level 3 gridded data available on a 0.25 × 0.25 degree grid. As compared to TROPOMI, OMI has similar measuring principle and overpass time, but lower spatial resolution (13 × 24 km2 in nadir). Therefore, OMI cannot capture the smallest spatial patterns near the emission sources but provides consistent NO2 observations since 2005. We limit the analysis of the spring–summer anomalies based on OMI data to the year 201520 to avoid the potential effect of earlier NO2 reductions due to cleaner technologies or other emission changes23,24.
OCO-2 data and XCO2 anomaly
Orbiting Carbon Observatory-2 (OCO-259) is NASA’s CO2 observing mission launched in 2014. Each 0.333 s, OCO-2 provides 8 soundings with spatial footprint of 1.29 × 2.25 km2 across a swath less than 10 km wide. The column-averaged dry air mole fraction of CO2 (XCO2) is retrieved. Since CO2 is a long-lived greenhouse gas, it is transported and accumulated in the atmosphere. To extract the anthropogenic enhancements from such a large background signal, we calculate the XCO2 anomalies with respect to the daily medians for each 10-degree latitude band and linearly interpolated to each OCO-2 data point60. The CO2 anomalies are gridded into a regular 1 × 1 degree grid and averaged over the period March-August of each year. The averaged maps are generally based on 5–10 days of observations for each grid cell within this six-month period. The XCO2 anomaly difference maps are calculated with respect to the year 2021 (Fig. 2), as well as the period 2019–2021 (TROPOMI period) and 2017–2021 (OCO-2 period) (Supplementary Fig. 13). When changing the reference period, the spatial patterns in the difference map remain similar. Here we use the OCO-2 version V11r lite files, which include bias correction and data screening. The OCO-2 data product is available from NASA’s Goddard Earth Sciences Data and Information Services Center (GES DISC, https://disc.gsfc.nasa.gov).
VIIRS active fire/thermal anomaly data
To identify the location of fires and thermal anomalies, we use the 375 m active fire data from the Visible Infrared Imaging Radiometer Suite (VIIRS), operating on-board the joint NASA/NOAA Suomi-National Polar-orbiting Partnership (S-NPP) satellite since October 2011, as available from NASA’s Fire Information for Resource Management System (FIRMS) archive. S-NPP VIIRS data are available twice a day at the equator with overpass times approximately 13:30 PM (ascending node) and 1:30 AM (descending node). The VIIRS fire detection algorithm is based on the 375 m middle and thermal infrared imagery data (I-bands). The product includes (among other variables) the time and coordinates of the fires as well as the corresponding fire radiative power (FRP, in MW). We used the active fire data product over Ukraine available in shapefile format from the FIRMS archive download: https://firms.modaps.eosdis.nasa.gov/download/.
Sentinel 2 false color (urban) product
To provide context about the land properties and to further analyze the distribution of fires, we use enhanced natural color and false color images from Sentinel-2 as available from the Sentinel Hub EO Browser (https://apps.sentinel-hub.com/eo-browser). Sentinel-2 is a wide-swath multi-spectral imaging mission that provides high-resolution (10 m, 20 m, and 60 m, depending on the wavelength) images for 13 spectral bands in the visible and infrared wavelengths. The mission is based on a constellation of two identical satellites, Sentinel-2A and Sentinel-2B, both flying on a sun-synchronous orbit. The images have been available since March 2018 as 100 × 100 km2 tiles, with a maximum revisit time of 5 days (using both Sentinel-2A and 2B satellites). The Sentinel-2 mission is designed to monitor surface properties such as vegetation, soil and water cover. The false color (urban) products are based on the composite of bands 12, 11 and 4. Vegetation is visible in shades of green, while constructed areas are represented by white or gray. Snow and ice appear as dark blue, and water as black or blue. Surfaces with elevated temperatures, such as forest fires, calderas of volcanoes, gas flaring over oil and gas facilities as well as hot smoke from industrial activities, saturate the image in medium IR channels and appear in shades of red or yellow.
All satellite data used in this work are publicly available through NASA and ESA data distribution hubs. The S5P-PAL NO2 dataset reprocessed until November 2021 is available at: https://data-portal.s5p-pal.com. Additionally, the offline (OFFL) NO2 products used to cover the period after November 2021 is available from the Sentinel-5P Pre-Operations Data Hub (https://s5phub.copernicus.eu/). The OMI NO2 L3 data and the OCO-2 data can be accessed from NASA's GES DISC (https://disc.gsfc.nasa.gov). NASA VIIRS Active Fire data (SHP format) data set is provided by LANCE FIRMS and it is available online at: https://earthdata.nasa.gov/active-fire-data. Sentinel-2 images are available from the Sentinel Hub EO Browser (https://apps.sentinel-hub.com/eo-browser).
Reuters, Russia occupies 20% of Ukraine's territory – Zelenskiy, 2 June 2022, https://www.reuters.com/world/europe/russia-occupies-20-ukraines-territory-zelenskiy-2022-06-02/.
Office of the high commissioner for human right (OHCHR), Ukraine: Civilian casualty update 13 February 2023, https://www.ohchr.org/en/news/2023/02/ukraine-civilian-casualty-update-13-february-2023.
UNHCR, Operational Data Portal, Ukraine refugee situation, https://data.unhcr.org/en/situations/ukraine.
World Bank. Ukraine Rapid Damage and Needs Assessment (English). Washington, D.C.: World Bank Group, 31 July 2022. http://documents.worldbank.org/curated/en/099445209072239810/P17884304837910630b9c6040ac12428d5c.
United Nations Environment Programme (UNEP), “The Environmental Impact of the Conflict in Ukraine: A Preliminary Review”. Nairobi, Kenya, https://wedocs.unep.org/20.500.11822/40746 (2022).
Pereira, P., Bašić, F., Bogunovic, I. & Barcelo, D. Russian–Ukrainian war impacts the total environment. Sci. Total Environ. 837, 155865. https://doi.org/10.1016/j.scitotenv.2022.15586 (2022).
Rawtani, D., Gupta, G., Khatri, N., Rao, P. K. & Hussain, C. M. Environmental damages due to war in Ukraine: A perspective. Sci. Total Environ. 850, 157932. https://doi.org/10.1016/j.scitotenv.2022.157932 (2022).
NASA Earth Observatory, Larger Wheat Harvest in Ukraine Than Expected, https://earthobservatory.nasa.gov/images/150590/larger-wheat-harvest-in-ukraine-than-expected (2022).
Lin, F. et al. The impact of Russia–Ukraine conflict on global food security. Glob. Food Sec. 36, 100661. https://doi.org/10.3390/foods11152301 (2023).
EcoZagroza, Official resource of the Ministry of Environmental Protection and Natural Resources of Ukraine, https://ecozagroza.gov.ua/en.
Conflict and Environment Observatory (CEO), Ukraine conflict environmental briefing, https://ceobs.org/ukraine-conflict-environmental-briefing-industry/.
de Klerk, L. et al. Climate damage caused by Russia’s war in Ukraine, 1 June 2023, https://en.ecoaction.org.ua/wp-content/uploads/2023/06/clim-damage-by-russia-war-12months.pdf.
Rajaeifar, M. et al. Decarbonize the military—mandate emissions reporting. Nature 611, 29–32. https://doi.org/10.1038/d41586-022-03444-7 (2022).
Krotkov, N. A. et al. Aura OMI observations of regional SO2 and NO2 pollution changes from 2005 to 2015. Atmos. Chem. Phys. 16, 4605–4629. https://doi.org/10.5194/acp-16-4605-2016 (2016).
Castellanos, P. & Boersma, K. F. Reductions in nitrogen oxides over Europe driven by environmental policy and economic recession. Sci. Rep. 2, 265. https://doi.org/10.1038/srep00265 (2012).
de Foy, B., Lu, Z. & Streets, D. Impacts of control strategies, the Great Recession and weekday variations on NO2 columns above North American cities. Atmos. Environ. 138, 74–86. https://doi.org/10.1016/j.atmosenv.2016.04.038 (2016).
de Ruyter de Wildt, M., Eskes, H. & Boersma, K. F. The global economic cycle and satellite-derived NO2 trends over shipping lanes. Geophys. Res. Lett. 39, L01802. https://doi.org/10.1029/2011GL049541 (2012).
Beirle, S. et al. Wagner, Pinpointing nitrogen oxide emissions from space. Sci. Adv. 5, eaax9800. https://doi.org/10.1126/sciadv.aax9800 (2019).
Beirle, S. et al. Catalog of NOx emissions from point sources as derived from the divergence of the NO2 flux for TROPOMI. Earth Syst. Sci. Data 13, 2995–3012. https://doi.org/10.5194/essd-13-2995-2021 (2021).
Liu, F. et al. Abrupt decline in tropospheric nitrogen dioxide over China after the outbreak of COVID-19. Sci. Adv. https://doi.org/10.1126/sciadv.abc2992 (2020).
Goldberg, D. L. et al. Disentangling the impact of the COVID-19 lockdowns on urban NO2 from natural variability. Geophys. Res. Lett. 47, e20200GL089269. https://doi.org/10.1029/2020GL089269 (2020).
Bauwens, M. et al. Impact of coronavirus outbreak on NO2 pollution assessed using TROPOMI and OMI observations. Geophys. Res. Lett. 47, e2020GL087978. https://doi.org/10.1029/2020GL087978 (2020).
Cooper, M. J. et al. Global fine-scale changes in ambient NO2 during COVID-19 lockdowns. Nature 601, 380–387. https://doi.org/10.1038/s41586-021-04229-0 (2022).
Georgoulias, A. K., van der A, R. J., Stammes, P., Boersma, K. F. & Eskes, H. J. Trends and trend reversal detection in 2 decades of tropospheric NO2 satellite observations. Atmos. Chem. Phys. 19, 6269–6294. https://doi.org/10.5194/acp-19-6269-2019 (2019).
Abeed, R. et al. A space view of agricultural and industrial changes during the Syrian civil war. Elementa Sci. Anthr. 9(1), 000041. https://doi.org/10.1525/elementa.2021.000041 (2021).
Janssens-Maenhout, G. et al. Toward an operational anthropogenic CO2 emissions monitoring and verification support capacity. Bull. Am. Meteorol. 101(8), E1439–E1451. https://doi.org/10.1175/BAMS-D-19-0017.1 (2020).
Reuter, M. et al. Towards monitoring localized CO2 emissions from space: Co-located regional CO2 and NO2 enhancements observed by the OCO-2 and S5P satellites. Atmos. Chem. Phys. 19, 9371–9383. https://doi.org/10.5194/acp-19-9371-2019 (2019).
Hakkarainen, J. et al. Analyzing nitrogen oxides to carbon dioxide emission ratios from space: A case study of Matimba Power Station in South Africa. Atmos. Environ. X 10, 100110. https://doi.org/10.1016/j.aeaoa.2021.100110 (2021).
Hakkarainen, J. et al. Building a bridge: Characterizing major anthropogenic point sources in the South African Highveld region using OCO-3 carbon dioxide snapshot area maps and Sentinel-5P/TROPOMI nitrogen dioxide column. Environ. Res. Lett. 18, 035003. https://doi.org/10.1088/1748-9326/acb837 (2023).
Veefkind, P. et al. TROPOMI on the ESA Sentinel-5 Precursor: A GMES mission for global observations of the atmospheric composition for climate, air quality and ozone layer applications. Remote Sens. Environ. 120, 70–83. https://doi.org/10.1016/j.rse.2011.09.027 (2012).
Kyiv School of Economics (KSE), Assessment of damages in Ukraine due to Russia's military aggression as of September 1, 2022, https://kse.ua/wp-content/uploads/2022/10/ENG-Sep22_Working_Sep1_Damages-Report.docx.pdf (2022).
NASA Earth Observatory, Tracking Night Lights in Ukraine, https://earthobservatory.nasa.gov/images/150002/tracking-night-lights-in-ukraine.
The New York Times (NYT), Ukraine Goes Dark: Images From Space Drive Home the Nation’s Anguish, 30 March 2023, https://www.nytimes.com/2023/03/30/world/europe/ukraine-satellite-darkness.html.
The Kyiv Independent, Klitschko: Kyiv’s population returns to pre-war level, 30 December 2022, https://kyivindependent.com/news-feed/klitschko-kyivs-population-returns-to-pre-war-level.
Reuters, Russian forces capture Ukraine’s second biggest power plant, Ukraine says, 28 July 2022, https://www.reuters.com/world/europe/fate-ukraines-second-biggest-power-plant-balance-after-russian-advance-2022-07-27/.
Epravda, Ukrenergo announced a critical situation with coal stocks in TPP warehouses (in Ukrainian), https://www.epravda.com.ua/news/2021/01/19/670142/.
International crisis group, Picturing the humanitarian response in western Ukraine, 28 July 2022, https://www.crisisgroup.org/europe-central-asia/eastern-europe/ukraine/picturing-humanitarian-response-western-ukraine.
Zhang, C., Hu, Q., Su, W., Xing, C. & Liu, C. Satellite spectroscopy reveals the atmospheric consequences of the 2022 Russia–Ukraine war. Sci. Total Environ. 869, 161759. https://doi.org/10.1016/j.scitotenv.2023.161759 (2023).
Independent, Ukraine-Russia: Kyiv mayor Vitali Klitschko imposes curfew as capital facing ‘dangerous moment’, 15 March 2022, https://www.independent.co.uk/news/world/europe/kyiv-curfew-ukraine-russia-invasion-b2036470.html.
Pro Consulting, Market Research Report on the Freight Transport Market (Railway and Road Transport) in Ukraine, https://pro-consulting.ua/ua (2022).
Trading Economics, Ukraine GDP Annual Growth Rate, https://tradingeconomics.com/ukraine/gdp-growth-annual. Accessed 1 Feb 2023.
Trading Economics, Ukraine Industrial Production, https://tradingeconomics.com/ukraine/industrial-production. Accessed 1 Feb 2023.
Grigorenko, Y. Industrial production in Ukraine in 2022, https://gmk.center/en/infographic/industrial-production-in-ukraine-decreased-by-37-in-2022/. Accessed 21 June 2023.
Ministry of Energy of Ukraine, https://mev.gov.ua/home. Last accessed 8 June 2023.
International Energy Agency, Ukraine Real-Time Electricity Data Explorer, https://www.iea.org/data-and-statistics/data-tools/ukraine-real-time-electricity-data-explorer.
Live Universal Awareness Map (Liveuamap), History of shelling over Ukraine, https://liveuamap.com.
CNN, Ukrainians have retaken Irpin from the Russian invaders, 31 March 2022, https://edition.cnn.com/2022/03/31/europe/irpin-ukraine-war-destruction-intl/index.html.
International Organization for Migration, Ukraine - Internal Displacement Report - General Population Survey Round 5 (17–23 May 2022), https://dtm.iom.int/reports/ukraine-internal-displacement-report-general-population-survey-round-5-17-23-may-2022.
United Nations Department of Economic and Social Affairs, Population Division, World Population Prospects 2022: Summary of Results. UN DESA/POP/2022/TR/NO. 3 (2022).
Román, M. O. et al. NASA’s Black Marble nighttime lights product suite. Remote Sens Environ. 210, 113–143. https://doi.org/10.1016/j.rse.2018.03.017 (2018).
Oda, T. & Maksyutov, S. A very high-resolution (1 km×1 km) global fossil fuel CO2 emission inventory derived using a point source database and satellite observations of nighttime lights. Atmos. Chem. Phys. 11, 543–556. https://doi.org/10.5194/acp-11-543-2011 (2011).
Meijer, Y. et al. Copernicus CO2 Monitoring Mission Requirements Document. Tech. rep., European Space Agency. Issue 3.0, EOP-SM/3088/YM-ym (2020).
Kasahara, M. et al. Overview and current status of GOSAT-GW mission and AMSR3 instrument. In Sensors, Systems, and Next-Generation Satellites XXIV, International Society for Optics and Photonics (eds Neeck, S. P., Hélière, A. & Kimura, T.) 1153007 (SPIE, 2020). https://doi.org/10.1117/12.2573914.
Ingmann, P. et al. Requirements for the GMES Atmosphere Service and ESA’s implementation concept: Sentinels-4/-5 and-5P. Remote Sens Environ. 120, 58–69. https://doi.org/10.1016/j.rse.2012.01.023.657 (2012).
Zoogman, P. et al. Tropospheric emissions: Monitoring of pollution (TEMPO). J. Quant. Spectrosc. Radiat. Transf. 186, 17–39. https://doi.org/10.1016/j.jqsrt.2016.05.008 (2017).
Kim, J. et al. New era of air quality monitoring from space: Geostationary environment monitoring spectrometer (GEMS). Bull. Am. Meteorol. Soc. 101, E1–E22. https://doi.org/10.1175/BAMS-D-18-0013.1 (2020).
S5P Product Algorithm Laboratory, 2021. S5P-PAL Data Portal. https://data-portal.s5p-pal.com. Accessed 16 May 2022.
Levelt, P. F. et al. The Ozone Monitoring Instrument: Overview of 14 years in space. Atmos. Chem. Phys. 18, 5699–5745. https://doi.org/10.5194/acp-18-5699-2018 (2018).
Crisp, D. et al. The on-orbit performance of the Orbiting Carbon Observatory-2 (OCO-2) instrument and its radiometrically calibrated products. Atmos. Meas. Tech. 10, 59–81. https://doi.org/10.5194/amt-10-59-2017 (2017).
Hakkarainen, J., Ialongo, I., Maksyutov, S. & Crisp, D. Analysis of four years of global XCO2 anomalies as seen by orbiting carbon observatory-2. Remote Sens. 11(7), 850. https://doi.org/10.3390/rs11070850 (2019).
State Statistics Service of Ukraine, Number of Present Population of Ukraine, as of January 1 2022, http://db.ukrcensus.gov.ua/PXWEB2007/ukr/publ_new1/2022/zb_Chuselnist.pdf (2022).
We thank the algorithm, processing, and distribution teams for the S5P/TROPOMI, OCO-2, OMI, Suomi-NPP/VIIRS and Sentinel 2 datasets used here. We also acknowledge the Ministry of Environmental Protection and Natural Resources of Ukraine for supporting the Finnish-Ukrainian research cooperation. Funding for this work was provided by the following projects: Ministry for Foreign Affairs of Finland, Institutional Cooperation Instrument (ICI), UHMC-FMI Meteorology project; Research Council of Finland ACCC grant 357904; Research Council of Finland CoE grant 353082.
The authors declare no competing interests.
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
About this article
Cite this article
Ialongo, I., Bun, R., Hakkarainen, J. et al. Satellites capture socioeconomic disruptions during the 2022 full-scale war in Ukraine. Sci Rep 13, 14954 (2023). https://doi.org/10.1038/s41598-023-42118-w