Thank you for visiting nature.com. You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.
Advertisement
Scientific Reports volume 16, Article number: 10437 (2026) Cite this article
1416 Accesses
Metrics details
Photovoltaic (PV) development in arid regions faces challenges such as sparse observational data, insufficient consideration of natural environmental heterogeneity, and a disconnect between site suitability assessments and actual power generation potential. To address these issues, this study integrates ERA5-Land reanalysis data, ESA CCI land cover data, DEM terrain data, and PV site information to construct a land suitability factor ranging from 0 to 1. Coupled with the Photovoltaic Library Python model(PVLIB-Python), a comprehensive assessment framework is established, spanning from site suitability to power generation potential and emission reduction benefits.Results show that Xinjiang’s theoretical PV generation potential from 2015 to 2025 averages approximately 113.5 PWh per year. After applying land suitability constraints, the technical potential decreases to 71.4 PWh annually, representing about 63% of the theoretical potential. Spatially, suitability follows a pattern of “concentration in basins and dispersion in mountainous areas,” with highly suitable zones mainly located in the central-western Tarim Basin, the Hami Basin, and the southern edge of the Junggar Basin.Based on the calculated technical potential, annual PV deployment could achieve around 53.5 billion tonnes of CO₂ emission reductions.Incorporating environmental benefits significantly lowers the levelized cost of energy (LCOE) for PV systems, demonstrating considerable net social value. This study provides quantitative evidence to support the scientific planning of PV power stations in Xinjiang and the formulation of carbon neutrality pathways.
As the global economy expands, the demand for fossil fuels continues to rise. This trend not only accelerates the depletion of finite resources but also exacerbates environmental degradation and global warming, primarily due to increased carbon dioxide (CO₂) emissions. According to the Sixth Assessment Report of the Intergovernmental Panel on Climate Change¹, achieving the Paris Agreement’s temperature goals requires a rapid and profound transformation of the global energy system. Renewable energy sources, particularly solar and wind power, are expected to play a decisive role in decarbonizing the electricity sector.Therefore, addressing energy shortages while advancing the transition to clean energy has become an urgent global challenge. Solar photovoltaic (PV) energy is widely recognized as a key renewable source capable of replacing fossil fuels in the future^2,3,4,5. Aligning with this direction, national policies encourage accelerating the planning and construction of large-scale PV projects in desert, Gobi, and arid regions, aiming to integrate economic growth with the green transition. Given its abundant solar resources, the development of PV power in Xinjiang holds strategic importance for achieving China’s carbon neutrality goals.
The global photovoltaic (PV) energy market is undergoing unprecedented rapid growth and has become a key driver of the energy transition. According to the International Renewable Energy Agency (IRENA), cumulative global PV capacity reached 1,419 gigawatts by the end of 2023⁶. By 2024, this capacity is projected to rise to 2,200 gigawatts, reflecting a worldwide shift toward renewable and sustainable energy technologies⁷. This expansion is driven by continuous technological advances, significant reductions in manufacturing costs, and strong global policy support—signaling a fundamental transformation in energy production and consumption patterns.As the world’s largest manufacturer and user of PV technology, China leads globally in both installed capacity and power generation. Within China, Xinjiang plays a pivotal role due to its abundant solar resources and vast available land.Beyond generating clean electricity, PV power plants also function as essential components within integrated hybrid energy systems. These systems often combine energy storage, wind power, diesel generators, and desalination facilities^8,9,10. Effective planning of such systems relies on accurate assessment of PV resources and suitable site selection, typically using methods like Geographic Information Systems (GIS) and multi-criteria decision analysis^11,12.Therefore, accurately evaluating the PV resource potential in Xinjiang forms the scientific basis for planning and constructing efficient, stable hybrid energy systems. This is of great significance for optimizing the regional energy structure and supporting the clean energy transition.In summary, current research exhibits three notable shortcomings: (1) Suitability assessments predominantly remain confined to spatial suitability classification, failing to conduct refined coupled evaluations with key resource constraints such as the power generation potential of photovoltaic systems¹³; (2) Most studies employ static or medium-to-low resolution data, overlooking the dynamic impact of microtopography and local climate on generation efficiency. Furthermore, understanding of the ecological impacts of PV installations exhibits regional limitations; recent research indicates that arid ecosystems demonstrate greater resilience to PV deployment, whereas humid regions are more vulnerable¹⁴; (3) Comprehensive assessments rarely systematically quantify the synergistic environmental and economic benefits of photovoltaic development.
To address the research gaps identified above, this study innovatively develops an integrated assessment framework that combines high-resolution meteorological data, multi-source geospatial data (land use, topography, ecological conservation areas), and a photovoltaic system physics simulation model. This framework enables a seamless evaluation from land suitability to power generation potential and further quantifies the emission reduction benefits of PV development. Additionally, by incorporating the levelized cost of electricity (LCOE) and the social cost of carbon—and drawing on methodologies for quantifying comprehensive benefits¹⁵—this study monetizes the environmental benefits, thereby enhancing the decision-support value of the assessment results.
To quantitatively assess the suitability of sites for centralized photovoltaic power stations in Xinjiang, this study developed a multi-constraint evaluation framework. The framework ultimately produced a spatially continuous “land suitability factor” layer, with values ranging from 0 to 1. This factor will serve as a key spatial constraint in subsequent calculations of technical generation potential (Fig. 2). The entire analytical process follows steps (Figs. 1).
This study selected slope gradient, land cover type, and ecological conservation zones as the key constraint factors for PV site selection. Land cover types were reclassified into five categories¹⁶ (Table 1), with bare ground and sparsely vegetated areas considered the most suitable, while ecologically valuable zones were excluded. To reflect the importance of infrastructure proximity in reducing development costs for large-scale PV projects¹⁷, “urban and built-up lands” were graded as relatively favorable within this classification system. Slope gradients were also categorized into five levels¹⁸ (Table 2), with suitability decreasing as slope steepness increases. Core ecological conservation areas were directly classified as unsuitable (assigned a value of 0).Factor weights were determined based on prior literature^19,20,21,22, assigning a weight of 0.64 to land cover and 0.36 to slope. Ecological conservation zones were treated as a veto factor, overriding other criteria where applicable.Finally, the normalized factor layers were integrated in ArcGIS using the Weighted Overlay tool to generate a land suitability layer. The suitability value SS for each pixel was calculated using the following formula:
where W denotes the weighting factor and S represents the standardized score for each constraint factor. For areas designated as nature reserves, all pixel values were set to zero, effectively excluding them from suitability consideration.To validate the model, the locations of existing photovoltaic power stations were overlaid with the generated suitability map. The model’s alignment with actual siting decisions was then assessed by calculating the proportion of stations located within each suitability zone²³.
Systematic framework for assessing the potential of power generation technologies; including input data and methods employed.
The following comparison of spatial distributions of CF values before and after condition constraints is presented. (a) Spatial distribution map of the calculated CF values, with the upper left corner displaying the probability density distribution of CF values; µ denotes the mean value of spatial CF values. (b) Spatial distribution map of CF values after conditional constraints, with the upper left corner showing the probability density distribution of CF values post-constraint; µ represents the mean value of spatial Corrected CF values; Value indicates the actual numerical value of CF.
To accurately assess the photovoltaic (PV) power generation potential in Xinjiang, this study utilizes a physics-based energy modeling framework for PV systems. The framework integrates high-resolution meteorological data with PV module technical parameters and employs the PVLIB-Python tool for systematic and reproducible performance simulations. PVLIB (Photovoltaic Library) is an open-source toolkit for modeling, simulating, and analyzing PV system performance. Widely recognized for its reliability and engineering applicability, it is developed and maintained by Sandia National Laboratories in the United States²⁴.
The assessment is primarily based on three core indicators: Capacity Factor (CF), Theoretical Potential, and Technical Potential. The capacity factor is defined as the ratio of a PV system’s actual electricity output to its theoretical maximum (rated) output over the same period, reflecting the system’s operational efficiency. Theoretical potential refers to the annual electricity generation based solely on available solar radiation, without considering land constraints. Technical potential represents the feasible generation capacity after applying spatial restrictions, such as land suitability, slope, and ecological conservation, to the theoretical potential¹⁹.
The direct current output power of photovoltaic modules²⁵ is primarily determined by the incident irradiance and the cell operating temperature. Its fundamental physical model is as follows:
In the equation,(:{E}_{PV}) represents the photovoltaic module’s direct current power (W), P_STC denotes the module power under standard test conditions (220 W), γ_pdc represents the power temperature coefficient (-0.0042/°C), T_cell indicates the cell operating temperature (°C), T_STC signifies the standard test temperature (25 °C), G_poa denotes the total irradiance (W/m²), and G_STC represents the standard test irradiance (1000 W/m²).    
The battery temperature is calculated using the PVsyst thermal model, a method widely used in engineering simulation²⁶, which accounts for the combined heating effects of ambient temperature and solar irradiance:
Where: T_amb denotes ambient temperature (°C), Uc denotes the conductive heat loss coefficient (29.0 W/m²·°C), Uv denotes the convective heat loss coefficient (W/m²·°C·(m/s)), V denotes wind speed (m/s).
This study employs PVLIB-Python (version 0.13.0) to simulate the hourly performance of photovoltaic systems across a 0.1° × 0.1° spatial grid. A standardised simplified full-coverage layout model is utilised to assess regional-scale theoretical and technical potential for photovoltaics. Given the geographical similarities between Xinjiang and the Qinghai-Tibet Plateau, we referenced literature on photovoltaic power generation in the latter region¹⁹. Canadian Solar CS5P 220 M modules(polycrystalline silicon) were selected, paired with ABB MICRO-0.25-I-OUTD-US 208Vac inverters (rated efficiency 96%). System configuration adheres to specified component parameters (Table 3). The photovoltaic array employs fixed mounting structures (38° tilt angle, 180° azimuth facing south) to optimise solar capture efficiency. The simulation follows a three-stage workflow(Fig. 1).
(1) Solar Geometry and Irradiance Calculation: The solar position algorithm computes hourly solar altitude and azimuth angles. Combined with the ERBS model, global horizontal irradiance (GHI) is decomposed into direct normal irradiance (DNI) and diffuse horizontal irradiance (DHI). The Perez model then calculates the total irradiance incident on the tilted plane (POA). (2) Temperature and Power Conversion: The PVsyst thermal model estimates cell operating temperature. Using the module’s temperature coefficient, the direct current (DC) output power is derived and then converted to alternating current (AC) output power via the inverter efficiency. (3)Performance and Potential Estimation: The capacity factor (CF) is calculated as the ratio of AC output to the rated DC power, indicating the system’s operational efficiency. The physical area of each grid cell (in km²) is calculated based on latitude. Assuming an installation density of 30 MW/km²²⁷ and using the derived CF, the annual theoretical power generation per grid cell is estimated with the following formula. Summing these values yields the total PV power generation potential for Xinjiang:
Here, ρ denotes the installed density (30 MW/km²), CFt is the capacity factor for the t^(th) hour of the year, and A is the area of each grid cell (km²).
To more comprehensively assess the competitiveness of photovoltaic power generation, this study extends the traditional techno-economic analysis framework by incorporating environmental externalities. The core of this approach involves monetizing the environmental benefits from avoided CO₂ emissions and integrating them into the levelized cost of energy (LCOE) calculation. This provides a more accurate reflection of its overall societal cost-benefit profile.
The levelized cost of energy (LCOE) is a key metric for evaluating the economic feasibility of energy projects. To account for the environmental value of photovoltaics, this study applies the following formula²⁸:
Where: LCOE denotes the levelised cost of energy including environmental costs (yuan/kWh). r represents the discount rate (%).n denotes the full life cycle of the photovoltaic system (years). C represents the initial total investment cost of the photovoltaic system (RMB). C_O&M denotes the annual operational and maintenance cost of the system (RMB/year) (calculated at 1.5% of the unit cost). C_CO2 represents the annual environmental damage cost avoided by photovoltaic power generation, i.e. its environmental benefit (RMB/year). Et denotes the annual electricity generation of the photovoltaic system (kWh/year).
The environmental benefits of photovoltaic power generation in terms of CO₂ emissions avoided by substituting fossil fuels are measured using the following formula²⁹:
EF_CO2: Carbon dioxide emission factor of the replaced grid (kg CO₂/kWh). ϕ_co2: Social cost of carbon, i.e. the marginal social damage caused by each unit of CO₂ emissions (yuan/ton CO₂). Following the value in Reference³⁰, this study sets it at USD 70/tonne CO₂.
Xinjiang (73°40′E–96°23′E, 34°25′N–49°10′N) covers a land area of 1,664,900 km². Its topography is characterized by the ‘three mountains flanking two basins’: the Altai Mountains to the north, the Kunlun and Tianshan ranges to the south, and the Junggar and Tarim Basins to the north and south of the Tianshan Mountains, respectively (Fig. 3). This vast area, with its abundant sunshine and low population density, offers unique advantages for developing large-scale photovoltaic power stations³¹. Located along the Silk Road Economic Belt, Xinjiang features a typical temperate continental climate, with landscapes largely covered by deserts, Gobi areas, and sparse vegetation. The region receives abundant solar radiation and exhibits moderate topographic variation, forming favourable natural conditions for photovoltaic power generation. These advantages—especially the high solar availability and extensive areas of low ecological sensitivity—make Xinjiang highly suitable for large-scale centralized PV stations. Over the past decade, the region has undergone rapid economic growth, with carbon emissions expected to continue increasing in the coming years³², further highlighting the importance of transitioning to renewable energy sources.
Land cover classification and spatial distribution of photovoltaic power stations in Xinjiang. The map illustrates five land use categories (defined by suitability for solar deployment), with existing PV stations marked by red diamonds.The inset map in the bottom-right corner shows the location of Xinjiang within China.
This study employs the ERA5-Land reanalysis dataset (provided by the European Centre for Medium-Range Weather Forecasts, ECMWF) as its meteorological input(https://cds.climate.copernicus.eu/). ERA5-Land is the high-resolution land component of the ERA5 global reanalysis, generated by combining global land surface model simulations with multi-source observational data assimilation. It offers high spatiotemporal consistency and reliability. With a spatial resolution of 0.1° × 0.1° and hourly continuous data from 1950 onward, the dataset is well-suited for high-precision surface process simulations and photovoltaic potential assessments.As a next-generation high-resolution land reanalysis product, ERA5-Land has demonstrated strong reliability in simulating various surface processes³³ and is widely used in studies on surface energy balance, hydrometeorology, and renewable energy. Its capability across diverse climates is supported by evaluations showing good long-term consistency, such as in a 70-year precipitation assessment for Spain³⁴.For this study, hourly ERA5-Land data from 2015 to 2025 were extracted for the Xinjiang region, resampled to a 0.1° grid, and used as input for the PVLIB-Python model. Selected variables are listed in Table 4.
This study employed multiple spatial datasets to support the analysis. The Digital Elevation Model (DEM) data, sourced from NASA and USGS with a spatial resolution of 30 m(https://search.earthdata.nasa.gov/), was processed in ArcGIS to derive slope information for topographic characterization in Xinjiang. Land cover data were obtained from the ESA Climate Change Initiative (CCI) project (https://maps.elie.ucl.ac.be/CCI/) at 300 m resolution and reclassified into five categories relevant to this study (Fig. 3). Additionally, data on ecological functional protection zones(https://www.resdc.cn/) were incorporated to identify areas unsuitable for photovoltaic development.
The sectoral carbon emissions dataset utilised in this study originates from the 2022 provincial emissions inventory released by the China Carbon Emission Accounts and Datasets (CEADs) platform (https://www.ceads.net.cn/). The CEADs database was established in 2016 under the leadership of Professor Guan Dabo’s team at Tsinghua University. This platform integrates multi-source high-resolution remote sensing data, including satellite imagery and thermal infrared observations, alongside multi-dimensional enterprise and provincial/municipal statistical records, with the objective of compiling refined carbon accounting inventories. To meet the research requirements, the original 47 sectors within the CEADs dataset were consolidated into six primary sectors. Additionally, to supplement global-scale emissions data, this study incorporates the Global Atmospheric Emissions Database for Global Regions (EDGAR) data shared by the European Commission (https://edgar.jrc.ec.europa.eu/). EDGAR employs independent emission estimation methodologies consistent with the Intergovernmental Panel on Climate Change (IPCC) and provides emission estimates for comparison with emissions reported by Parties to the United Nations Framework Convention on Climate Change (UNFCCC). This database features a spatial resolution of 0.1° × 0.1°. This study selected EDGAR’s 2022 fossil CO₂ emissions data, which encompasses all fossil CO₂ sources. These include combustion of fossil fuels, processing of non-metallic minerals (such as cement production), metal production processes (involving both ferrous and non-ferrous compounds), urea production, agricultural lime use, and solvent utilisation across various industrial and agricultural activities.
In order to validate the reliability and accuracy of the site suitability model, this study employed existing photovoltaic location data from Xinjiang²³ to assess the suitability framework developed herein. The Zhang et al. dataset features a spatial resolution of 30 m, utilising Landsat series remote sensing imagery as foundational data. Constructed through a random forest algorithm and the Google Earth Engine platform, it achieves high-precision identification and mapping of photovoltaic power stations across China. Its classification accuracy has been validated through field verification and cross-validation with high-resolution imagery, demonstrating considerable reliability and authority. The dataset constructed by Feng et al. features a spatial resolution of 10 m, utilising Sentinel-2 remote sensing imagery as its foundational data. Employing a random forest algorithm and active learning strategy, it achieves high-precision identification and mapping of ground-mounted photovoltaic power stations across China in 2020. Its classification accuracy, verified to reach 89%, makes it the first nationwide open-source photovoltaic power station map at 10-metre resolution, possessing significant application value.
Simulations using the PVLIB-Python model and ERA5-Land data show that Xinjiang has substantial theoretical photovoltaic (PV) generation potential, though technical feasibility is significantly constrained by land availability. From 2015 to 2025, the average annual theoretical generation potential was approximately 113.5 PWh. After applying land suitability constraints, the technical potential decreased to 71.4 PWh, representing about 63% of the theoretical value.Spatially, both solar radiation and the capacity factor (CF) follow a “high in basins, low in mountainous areas” pattern (Fig. 4a,b). High-value zones are concentrated in the Hami Basin, eastern Tarim Basin, and southwestern Junggar Basin, with a spatially averaged CF of about 0.167. Lower values are mainly found on the northern slopes of the Tianshan Mountains, the southern foothills of the Altai Mountains, and the oasis belt along the Tarim River, reflecting the influence of topography and vegetation cover.Temporally, both annual total radiation and annual mean CF show a slight upward trend, with interannual fluctuations below 5%, indicating relatively stable solar resources and PV generation efficiency in the region over the study period (Fig. 4c,d).
Spatial patterns of CF in Xinjiang from 2015 to 2025. (a) Spatial distribution of mean annual total solar radiation in Xinjiang, 2015–2025. The upper left panel shows the probability density distribution of annual mean total radiation, with “mean” indicating the spatial average of annual total radiation. (b) Spatial distribution of conditionally constrained CF values. The upper left panel shows the probability density distribution of CF, with “mean” indicating the spatial average of CF. (c) Temporal trend of total solar radiation in Xinjiang from 2015 to 2025. (d) Probability density of the temporal trend in average CF for Xinjiang from 2015 to 2025.
The PV suitability map for Xinjiang, generated based on constraints including slope, land cover type, and ecological conservation zones (Fig. 5a), exhibits a clear spatial distribution of “concentration in basins and dispersion in mountainous areas.”Highly suitable zones are mainly concentrated in the central-western Tarim Basin, the entire Hami Basin, and the southern margin of the Junggar Basin. These areas are predominantly bare land or sparsely vegetated (Table 1), with slopes generally between 0–6° (Table 2), and are located outside of ecological protection zones. This combination offers low ecological constraints and high construction feasibility.Moderately suitable areas are distributed around the peripheries of highly suitable zones and within the western Ili River valley. Some of these areas have light vegetation cover or gentle slopes and would require localized engineering measures—such as vegetation clearing or slope modification—to reduce development difficulties.Unsuitable zones are largely located within the Bogda Peak Nature Reserve on the northern slopes of the Tianshan Mountains, the Kanas Water Conservation Area in the Altai Mountains, and the oasis belt along the Tarim River. These regions either fall within ecological protection redlines (suitability factor = 0) or feature slopes exceeding 15° combined with land cover such as water bodies and wetlands (Table 1), resulting in high ecological sensitivity and prohibitively high construction costs.
Validation results (Fig. 5) show that the spatial overlay of existing PV power stations with the suitability map confirms a significant agreement between model outputs and actual siting decisions(nearly 60% in high-suitability zones (≥ 0.6)). This outcome demonstrates the value of integrating land cover, topography, and ecological constraints.Nevertheless, 14.6% of the existing PV stations within the study area are situated in zones classified as unsuitable. Some of these are located in ecologically sensitive regions, such as the northern slopes of the Tianshan Mountains, the southern foothills of the Altai Mountains, and the oasis belt along the Tarim River (Fig. 5a). Others are found in or near urban areas (Fig. 5c). This discrepancy likely arises from differences between the theoretical constraints applied in this study and actual development conditions¹⁷, which in our model are treated as unsuitable³⁵.
Suitability distribution for photovoltaic site selection in Xinjiang. (a) Land use suitability map (0–1 scale), with existing PV sites marked by red diamonds. The inset in the upper left corner is a probability density plot of the suitability values, where “Mean” indicates the average suitability. (b) Statistical chart showing the percentage distribution of existing photovoltaic sites by suitability rating (0, 0–0.2, 0.2–0.4, 0.4–0.6, 0.6–0.8, 0.8–1) relative to the total number of sites. (c) Box-whisker plot illustrating the suitability values across different land cover types found at existing power stations. In this plot, the red line represents the mean, the green dashed line denotes the median, and the blue line indicates the standard deviation.
To quantify the environmental benefits of PV power generation, this study assesses its emission reduction potential based on the following methodology and boundaries: (1) Substitution Boundary: All PV generation is assumed to be integrated into the Northwest China regional grid. Emission reductions are calculated using the grid’s 2020 baseline emission factor (0.749 kg CO₂/kWh), representing the average emissions displaced per unit of PV electricity.(2)Spatial Boundary: Following the “production-side” accounting principle, all emission reduction benefits are attributed to Xinjiang (the location of electricity production), regardless of where the power is transmitted or consumed. (3)Temporal & Operational Boundary: Calculations are based on the theoretical annual generation potential. Losses from plant auxiliary power or potential grid curtailment are not deducted, reflecting the theoretical maximum resource development potential.
Based on the above parameters, and incorporating Xinjiang’s PV technical potential of 71.4 PWh along with relevant SO₂ and NO_x emission factors (0.187 g/kWh and 0.195 g/kWh, respectively)¹⁹, the average annual CO₂ reduction from the region’s PV development is estimated at approximately 53.5 billion tonnes. Corresponding reductions in SO₂ and NO_x emissions reach about 13 Mt and 14 Mt per year, respectively.Spatially, carbon reduction benefits vary considerably (Fig. 6). In aggregate, the estimated reductions could offset roughly 100 times Xinjiang’s total carbon emissions in 2022. At a local level, certain prefectures and cities—such as Hotan Prefecture and Hami City—show offset potentials tens to hundreds of times greater than their 2022 emissions, far exceeding 100%. This is primarily because these areas coincide with zones of very high PV suitability and generation potential, where the emissions avoided through PV generation already surpass local carbon outputs.These results highlight Xinjiang’s significant role as a national clean energy base, capable of delivering substantial cross-regional carbon reduction benefits to the broader power system, beyond local emission cuts. It should be noted, however, that these high offset rates are derived from theoretical potential and fixed grid emission factors. In practice, factors such as real‑time grid absorption capacity and changing marginal emission factors would result in lower achievable reductions.
Spatial distribution of carbon reduction potential from photovoltaic power generation. The main map shows the estimated annual carbon reduction potential for each prefecture as a percentage of its own total carbon emissions in 2022. The donut chart (top-left) displays the carbon emission offset rate for each of Xinjiang’s five major industrial sectors relative to their respective 2022 emissions.
Under a baseline carbon price of US$70 per tonne of CO₂, the levelized cost of energy (LCOE) for photovoltaic systems, when environmental benefits are included, is -0.151 yuan/kWh. Without considering these benefits, the LCOE rises to 0.211 yuan/kWh. The monetized value of the environmental benefits is approximately 0.362 yuan/kWh, demonstrating that PV development provides substantial net social value (Fig. 7a). The LCOE including environmental benefits decreases linearly with a rising carbon price, declining by about 0.054 yuan/kWh for every US$10 per tonne increase. The breakeven carbon price—where LCOE equals zero—is approximately US$40.8 per tonne. Since the baseline price exceeds this threshold, it further confirms the economic sustainability of PV development (Fig. 7b).
Sensitivity analysis of levelised cost of energy (LCOE) to carbon social costs under Xinjiang’s theoretical photovoltaic power generation potential. (a) Relationship where LCOE decreases with rising carbon prices: the comparison between the solid green line (incorporating environmental benefits) and the dashed red line (excluding environmental benefits) clearly demonstrates the contribution of carbon emission reduction value towards lowering societal costs; When the carbon price exceeds the critical threshold (red marker), LCOE falls below the zero-cost line (black line), indicating that photovoltaic power generation generates net societal benefits. The baseline scenario of this study (US$70/tonne, orange marker) falls within this region. (b) Quantifies the specific contribution of environmental benefits to reducing LCOE (blue curve), which increases linearly with carbon price.
A sensitivity analysis of key PV system design parameters reveals that installation density and tilt angle distinctly influence power generation performance³⁶. Within the range of 20–40 MW/km², changes in installation density do not affect the average capacity factor, which remains stable at approximately 0.263. However, total theoretical power generation increases linearly with installation density(Fig. 8a). As the tilt angle increases from 33° to 44°, both the average capacity factor and total theoretical power generation gradually increase, with the rate of increase slowing around 38° (Fig. 8b). The results indicate that installation density mainly constrains the scale of power generation, while tilt angle simultaneously influences both system efficiency and power output. The values used in this study—30 MW/km² for installation density and 38° for tilt angle—represent a reasonable balance between efficiency and scale.
Sensitivity analysis of PV system installation parameters on power generation performance. (a) Shows the impact of installation density on average capacity factor (blue circular markers) and total theoretical generation (green square markers). (b) Illustrates the effect of tilt angle on both metrics. The large red dots denote the baseline conditions of 30 MW/km² installation density and 38° tilt angle. The left vertical axis (blue axis) denotes the average capacity factor, corresponding to the blue circular markers and curve in the figure. The right vertical axis (green axis) denotes the total theoretical power generation, corresponding to the green square markers and curve in the figure.
This study has several limitations. First, while it incorporates constraints such as land use type and ecological conservation zones in assessing site suitability and generation potential—emphasizing the theoretical upper limit of resources and technology—it does not account for additional external factors. These include current grid absorption capacity, infrastructure development, climate and soil conditions, and socio-economic influences. As a result, the assessment focuses primarily on revealing the region’s objective resource potential, which explains why the estimated technical potential is notably higher than values reported in some existing studies^13,31. Second, although constraints like land suitability and ecological red lines are included, the study does not quantitatively analyze potential local ecological and environmental impacts of PV construction—such as land occupation, local pollutants, or noise. It also does not systematically incorporate dynamic socio-technical factors like water resource availability, grid capacity, or economic cost fluctuations³⁷. Therefore, there remains room to enhance the comprehensiveness of planning support offered by this research.
To improve the scientific rigor and practical relevance of photovoltaic (PV) potential assessments, future work should focus on establishing open and standardized data platforms and evaluation workflows. These should integrate multi-source remote sensing and ground-based observation data, applying machine learning methods to minimize subjective bias. At the same time, a nationally unified data standard should be developed systematically^23,38 to support the optimized deployment of renewable energy.Furthermore, large-scale PV deployment brings not only emission reduction benefits but also socio-environmental implications. There is a need to develop dynamic, coupled assessment models that incorporate multi-dimensional constraints, such as resource potential, techno-economic feasibility, grid capacity, ecological conditions, and social acceptance. Such models would help align the energy transition with regional sustainable development in a synergistic manner.
This study integrates geographic constraints, meteorological data, and photovoltaic (PV) modeling to assess the suitability and generation potential of PV power in Xinjiang. Results show that PV site suitability follows a pattern of “concentration in basins and dispersion in mountainous areas.” Highly suitable zones are mainly located in the central-western Tarim Basin, the Hami Basin, and the southern margin of the Junggar Basin, characterized by bare or sparsely vegetated land, gentle slopes, and an absence of core ecological conservation areas.Simulations indicate that the theoretical annual PV generation potential in Xinjiang is approximately 113.5 PWh. After applying land suitability constraints, the technical potential decreases to about 71.4 PWh—roughly 63% of the theoretical value—highlighting land availability as a key limiting factor.Based on the technical potential, PV systems could achieve annual CO₂ emission reductions of approximately 53.5 billion tonnes, demonstrating considerable cross-regional carbon mitigation benefits. Incorporating the social cost of carbon further improves the economic viability of PV projects, underscoring the importance of environmental benefits in energy decision-making.This study provides methodological guidance for PV resource assessment in arid regions, with findings that can support PV planning in Xinjiang. Future work should integrate additional constraints, such as grid absorption capacity, to improve the systematic feasibility of energy planning.
The data code used in this study is available from the corresponding author on request.
Alternating current
Analytic hierarchy process
Capacity factor
Carbon emission accounts and datasets
Concentrated solar power
Direct current
Digital elevation model
Diffuse horizontal irradiance
Direct normal irradiance
Emissions database for global atmospheric research
European Centre for Medium-Range Weather Forecasts
European Space Agency Climate Change Initiative
Global horizontal irradiance
Geographic information system
International energy agency
Intergovernmental Panel on Climate Change.
International Renewable Energy Agency
Levelized cost of energy
National Aeronautics and Space Administration.
Plane of array
Photovoltaic
Photovoltaic library
United Nations Framework Convention on Climate Change
United States Geological Survey
IPCC. Climate Change 2022: Mitigation of Climate Change. Contribution of Working Group III to the Sixth Assessment Report of the Intergovernmental Panel on Climate Change (Cambridge University Press, 2022).
Geweda, A. E., Zayed, M. E., Khan, M. Y. & Alquaity, A. B. Mitigating CO2 emissions: A review on emerging technologies/strategies for CO₂ capture. J. Energy Inst. 118, 101911 (2025).
Article  CAS  Google Scholar 
Qiao, B. et al. Low-carbon dispatch of seasonal park integrated energy system based on adjustable cooling–heating–power ratio. Energies 18(19), 5071 (2025).
Abdelkareem, M. A. et al. Environmental aspects of fuel cells: A review. Sci. Total Environ. 752, 141803 (2021).
Article  CAS  PubMed  Google Scholar 
Asongu, S. A., Agboola, M. O., Alola, A. A. & Bekun, F. V. The criticality of growth, urbanization, electricity and fossil fuel consumption to environment sustainability in Africa. Sci. Total Environ. 712, 136376 (2020).
Article  CAS  PubMed  Google Scholar 
IEA. Integrating Solar and Wind (IEA, 2024).
Nassar, Y. F. et al. Technical and environmental cost-benefit analysis of strategies towards a green economy in the electricity sector in Libya. Economics Policy Energy Environment (2025).
Abduallah, A., Nassar, Y., Mohammed, S. & Ghatus, M. Integrating electricity sub-grid with pumped hydropower storage system for grid stability and sustainability. Wadi Alshatti Univ. J. Pure Appl. Sci. 322–332 (2025).
Nassar, Y. F. et al. Dynamic analysis and sizing optimization of a pumped hydroelectric storage-integrated hybrid PV/Wind system: A case study. Energy. Conv. Manag. 229, 113744 (2021).
Article  Google Scholar 
Nassar, Y. F. et al. Design of an isolated renewable hybrid energy system: A case study. Mater. Renew. Sustain. Energy 11(3), 225–240 (2022).
Rasaei, F. et al. Optimal selection of CSP site for desalination system using GIS and AHP method in Hormozgan province. Iran. Energy Reports. 13, 2255–2268 (2025).
Article  Google Scholar 
Abdoos, M. et al. Energy PLAN-based evaluation of renewable energy scenarios in semi-arid contexts. Results Eng. 107007 (2025).
Zhang, Y., Ren, J., Pu, Y. & Wang, P. Solar energy potential assessment: A framework to integrate geographic, technological, and economic indices. Renew. Energy. 149, 577–586 (2020).
Article  Google Scholar 
Xiao, J. et al. Ecological dichotomies of solar energy expansion: resilience in arid regions versus fragility in humid ecosystems. Front. Plant Sci. 16, 1549519 (2025).
Article  PubMed  PubMed Central  Google Scholar 
Lyu, F. et al. Quantifying the potential triple benefits of photovoltaic energy development in reducing emissions, restoring ecological resource, and alleviating poverty in China. Resour. Conserv. Recycl. 215, 108110 (2025).
Liu, L. et al. Potential contributions of wind and solar power to China’s carbon neutrality. Resour. Conserv. Recycl. 180, 106155 (2022).
Psarra, I. et al. Towards an integrated positive energy district landscape: technical and financial analysis for a case study in Hoogkerk. North. Neth. Energy Sustain. Soc. 15(1), 30 (2025).
Article  Google Scholar 
Mallick, J. Municipal solid waste landfill site selection based on fuzzy-AHP and geoinformation techniques in Asir Region, Saudi Arabia. Sustainability. 13(3), 1538 (2021).
Tang, W. et al. Meeting China’s electricity demand with renewable energy over Tibetan Plateau. Sci. Bull. 68(1), 39–42 (2023).
Rios, R. & Duarte, S. Selection of ideal sites for the development of large-scale solar photovoltaic projects through Analytical Hierarchical Process–Geographic information systems (AHP-GIS) in Peru. Renew. Sustain. Energy Rev. 149, 111310 (2021).
Article  Google Scholar 
Maheen, M. & Romshoo, S. A. Field-validated apple orchard land suitability mapping in Kashmir Valley using AHP-GIS and high-resolution satellite data. Spat. Inform. Res. 33(6), 1–19 (2025).
Google Scholar 
AbdelRahman, M. A., Yossif, T. M. & Metwaly, M. M. Enhancing land suitability assessment through integration of AHP and GIS-based for efficient agricultural planning in arid regions. Sci. Rep. 15(1), 31370 (2025).
Article  CAS  PubMed  PubMed Central  Google Scholar 
Zhang, X., Xu, M., Wang, S., Huang, Y. & Xie, Z. Mapping photovoltaic power plants in China using Landsat, random forest, and Google Earth Engine. Earth Syst. Sci. Data. 14(8), 3743–3755 (2022).
Article  ADS  Google Scholar 
Stein, J. S. The photovoltaic performance modeling collaborative (PVPMC). In 2012 38th ieee photovoltaic specialists conference 003048–003052 (IEEE, 2012).
Nassar, Y. F., Alsadi, S. Y., Ali, K. A., Yousef, A. H. & Fakher, M. A. Numerical analysis and optimization of area contribution of the PV cells in the PV/T flat-plate solar air heating collector. J. Solar Energy Res. Updates. 6, 43–50 (2019).
Google Scholar 
Barykina, E. & Hammer, A. Modeling of photovoltaic module temperature using Faiman model: Sensitivity analysis for different climates. Sol. Energy. 146, 401–416 (2017).
Article  ADS  Google Scholar 
He, G. & Kammen, D. M. Where, when and how much solar is available? A provincial-scale solar resource assessment for China. Renew. Energy. 85, 74–82 (2016).
Article  Google Scholar 
El-Khozondar, H. et al. Economic and environmental implications of solar energy street lighting in urban regions: A case study. Wadi Alshatti Univ. J. Pure Appl. Sci. 142–151 (2025).
Inweer, M. I. & Nassar, Y. Carbon emissions life cycle assessment of cement industry in Libya. Wadi Alshatti Univ. J. Pure Appl. Sci. 162–173 (2025).
Almhdi, E. & Miskeen, G. Power and carbon footprint evaluation and optimization in transitioning data centres. Wadi Alshatti Univ. J. Pure Appl. Sci. 221–229 (2025).
Bao, J. et al. Are regions conducive to photovoltaic power generation demonstrating significant potential for harnessing solar energy via photovoltaic systems?. Sustainability 16(8), 3281 (2024).
Google Scholar 
Yang, S. et al. Study on the carbon emission offset effect of vegetation in typical arid areas of Xinjiang, China. Ecol. Indic. 178, 114125 (2025).
Muñoz-Sabater, J. et al. ERA5-Land: A state-of-the-art global reanalysis dataset for land applications. Earth Syst. Sci. Data 13(9), 4349–4383 (2021).
Gomis-Cebolla, J., Rattayova, V., Salazar-Galán, S. & Francés, F. Evaluation of ERA5 and ERA5-Land reanalysis precipitation datasets over Spain (1951–2020). Atmos. Res. 284, 106606 (2023).
Article  Google Scholar 
Liang, X. et al. Impacts of land management practice strategy on regional ecosystems: Enlightenment from ecological redline adjustment in Jiangsu, China. Land Use Policy 119, 106137 (2022).
Ahmad, S., Agrira, A. & Nassar, Y. The impact of loss of power supply probability on design and performance of wind/pumped hydropower energy storage hybrid system. Wadi Alshatti Univ. J. Pure Appl. Sci. 52–62 (2025).
Al-Maghalseh, M. The environmental impact and societal conditions of PV power plants: A case study of Jericho Gate-Palestine Stat Of. Wadi Alshatti Univ. J. Pure Appl. Sci. 16–31 (2025).
Feng, Q. et al. A 10-m national-scale map of ground-mounted photovoltaic power stations in China of 2020. Sci. Data 11(1), 198 (2024).
Download references
This research was supported by the Basic Research Projects for Universities in the Autonomous Region (Grant XJEDU2025P007), National Natural Science Foundation of China (Grant 42461052), Tianshan Talents of the Autonomous Region (Third Batch) – Outstanding Young Talents – Young Scientific and Technological Innovation Talents(Grant 2024TSYCCX0025).
This research was supported by the Basic Research Projects for Universities in the Autonomous Region (Grant XJEDU2025P007).
College of Geography and Remote Sensing Sciences, Xinjiang University, Urumqi, 830046, China
Nuo Li, Wenjie Yu, Kexin Liu, Hanlu Zhang & Haiwei Zhang
Xinjiang Key Laboratory of Oasis Ecology, Urumqi, 830046, China
Haiwei Zhang
School of Intelligence Science and Technology, Xinjiang University, Urumqi, 830046, China
Yihang Zhou
Search author on:PubMed Google Scholar
Search author on:PubMed Google Scholar
Search author on:PubMed Google Scholar
Search author on:PubMed Google Scholar
Search author on:PubMed Google Scholar
Search author on:PubMed Google Scholar
Conceptualization: Kexin Liu; Methodology: Yihang Zhou; Formal analysis: Hanlu Zhang; Investigation: ;Writing—original draft preparation: Nuo Li and Haiwei Zhang; Writing—review and editing: Haiwei Zhang and Wenjie Yu; All authors have read and agreed to the published version of the manuscript.
Correspondence to Wenjie Yu or Haiwei Zhang.
The authors declare no competing interests.
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/.
Reprints and permissions
Li, N., Yu, W., Liu, K. et al. Reveal the deployable solar energy potential and emission reduction benefits in the arid areas of Xinjiang. Sci Rep 16, 10437 (2026). https://doi.org/10.1038/s41598-026-40841-8
Download citation
Received: 24 November 2025
Accepted: 16 February 2026
Published: 25 February 2026
Version of record: 27 March 2026
DOI: https://doi.org/10.1038/s41598-026-40841-8
Anyone you share the following link with will be able to read this content:
Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative
Collection
Advertisement
Scientific Reports (Sci Rep)
ISSN 2045-2322 (online)
© 2026 Springer Nature Limited
Sign up for the Nature Briefing: Anthropocene newsletter — what matters in anthropocene research, free to your inbox weekly.

source