Advanced photovoltaic technology can reduce land requirements and climate impact on energy generation

Solar PV is the third-largest renewable energy source globally and has grown rapidly, with recent deployment rates consistent with pathways to net-zero by mid-century. However, future PV deployment must contend with land requirements and climate change impacts that alter solar radiation and temperature, potentially reducing PV potential and increasing the land area needed to meet energy demand. The persistence of installed capacity over decades and uncertain rates of technological improvement add further complexity. Prior studies report mixed climate impacts across PV system types and regions and often neglect the interacting roles of technology advances and land-use competition at global and regional scales. This study addresses three questions: (1) How much land is required to meet future energy demand using conventional vs advanced PV technologies? (2) How do required land areas change when direct climate impacts on PV output are included? (3) How does PV potential vary when combined with other land uses?

Previous work using global circulation models suggests small global-scale changes in PV potential (e.g., up to −0.4% for large-scale PV, but +2% for rooftop PV by 2100), while localized changes at existing installations range from −19% to +16%. Regional model studies have projected substantial declines (up to 20% in South Asia and Latin America and up to 34% in Sweden). Land competition is frequently cited as limiting large-scale PV, though multifunctional uses (agrivoltaics over pastures, PV on highways, car parks, canals, and rooftops) can reduce land conflicts. Prior global assessments have rarely integrated climate impacts, technological advances across multiple PV types, and land requirements consistently across regions and scenarios. This study builds on recent PV potential modeling and SSP–RCP scenario frameworks to jointly evaluate climate, technology, and land-use interactions.

- Scenarios and climate data: Used four Representative Concentration Pathways (RCP2.6, RCP4.5, RCP6.0, RCP8.5) from four CMIP5 GCMs (HadGEM2-ES, MIROC5, GFDL-ESM2M, IPSL-CM5A-LR) with ISIMIP2b bias adjustment. Baseline period: 1991–2005 (WMO-consistent). Analyses target 2050 (2031–2070) and 2085 (2071–2100). - Energy demand scenarios: Global and regional energy demands from SSPs (SSP1, SSP2, SSP3, SSP4, SSP5) from IIASA; additional downscaled total energy projections used for regional comparisons. - PV technologies: Considered six PV technologies with efficiency ranges (details in Supplementary Table 1), focusing on conventional crystalline Si, perovskites, III–V cells, multijunctions with c-Si, and III–V multijunctions. Used mean technology efficiencies to compute outputs and land needs. Ground coverage ratio (GCR) standard value 0.51 used in sensitivity of area requirements. - PV potential calculation: Computed PV energy yield (kWh/kWp) using gridded solar radiation and air temperature; temperature effects applied using standard coefficients (efficiency reduces ~0.4–0.5% per °C above 25°C). Other losses (e.g., soiling) included via typical values; wind speed omitted due to large uncertainties in near-surface projections. - Sensitivity analysis: Univariate sensitivity to radiation and temperature separately and combined, averaged across GCMs and RCPs along decadal time series 2030–2100. - Land requirement estimation: For each region and globally, required PV land area (%) = energy demand / total PV energetic output over land for each technology and scenario. Permanent water and ice excluded; other siting suitability constraints not modeled. - Multifunctional land-use assessment: Explored PV potential on selected land categories and fractions: Sahara Desert (~7% of global land), highways (~0.7% of land), urban areas (projected 0.8–2.6% of land by 2100), and 2–5% of global grasslands (which cover ~20–40% of land). Compared potential PV generation to scenario demands. - Regionalization: Results reported for World Bank regions (EAP, ECA, LAC, MENA, NA, SA, SSA) with land areas and baseline PV potential as summarized in the article’s tables.

- Global land needs without climate impacts (2085): Conventional Si PV would require ~0.5–1.2% of global land to meet projected energy demands across SSPs; advanced technologies (e.g., perovskites, III–V multijunctions) would reduce this to ~0.3–1.0% (about one-half to three-quarters of Si area). - With climate impacts included (2085): Required land increases to ~0.7–1.5% for conventional Si. Advanced technologies can reduce this to ~0.3–1.2%, largely offsetting climate-related declines. - Climate drivers and PV potential: Despite small global increases in radiation (+0.4% to +1.0% by 2085), higher temperatures (+1.5°C to +5.0°C by 2085) reduce module efficiency, yielding global PV potential declines of −3.4% to −5.0% by 2085 (−3.7% to −4.5% by 2050). Efficiency loss is ~0.4–0.5% per °C above 25°C. - Regional climate impacts by 2085: PV potential declines up to ~−3% in Latin America and the Caribbean and up to ~−8% in South Asia, with strong geographic variability. - Typical PV yields: With typical losses, global mean PV potential is ~1340–1363 kWh/kWp. Regional extremes: Sub-Saharan Africa ~1621–1661 kWh/kWp (highest) and Europe & Central Asia ~976–989 kWh/kWp (lowest). - Area needs with GCR considered: Global land required becomes larger when accounting for GCR=0.51: Si ~1.0–2.4%, perovskite ~0.8–1.8%, III–V multijunction ~0.5–1.1%. - High-emissions case (SSP5–RCP8.5, 2085): Conventional Si requires ~1.5% of global land (or ~3% with GCR). III–V multijunction requires ~0.7% (or ~1.4% with GCR) to meet demand. - Regional area increases with climate: Additional land required by climate impacts is up to ~0.2–0.5 percentage points in South Asia and East Asia & Pacific, ~0.3–0.7 pp in Middle East & North Africa, and ~0.02–0.04 pp in Sub-Saharan Africa. - Multifunctional land-use potential: Sahara (~7% of land) could, hypothetically, supply ~5–11× global energy demand in 2085. Highways (~0.7% of land) could supply ~1.2× 2085 demand under SSP1–RCP2.6. Urban PV (0.8–2.6% land by 2100) could supply ~1.4–4.2× 2085 demand. Using 2% of global grasslands (0.4–0.8% of land) as intensive pasture agrivoltaics could supply ~0.3–1.3× global demand depending on technology and siting. - Regional opportunities: Sub-Saharan Africa shows especially high potential (could generate multiples of regional demand on modest land shares), with MENA also able to meet a substantial share of regional demand depending on technology and scenario.

The study shows that while climate change reduces PV potential primarily via temperature-driven efficiency losses, improvements in PV technologies more than compensate, thereby stabilizing or reducing land requirements to meet future energy demand. This addresses the research questions by quantifying land needs across technologies and scenarios, demonstrating the incremental land burden from climate impacts, and identifying how multifunctional deployments can alleviate land-use competition. Regional differences are pronounced, with larger climate penalties in South Asia and MENA, but high irradiance regions (e.g., Sub-Saharan Africa) retain strong potential. The results reinforce that technology innovation, rapid roll-out, and strategic siting can maintain PV’s capacity to meet demand under climate change. The discussion also notes system-level considerations (storage, grid integration), alternative configurations (floating PV), and environmental trade-offs of large-scale deployment, emphasizing the importance of context-specific planning and policy support.

Large-scale PV can meet future global energy demands even under climate change if advanced, higher-efficiency technologies are deployed and siting leverages multifunctional land uses. Climate change decreases PV potential by a few percent globally, increasing land needs modestly, but technology advances can offset these effects and reduce area requirements relative to conventional Si. Strategic deployment on urban infrastructure, highways, and agrivoltaic systems can further ease land competition. Future research should refine technology-specific siting under local climates, integrate detailed land suitability and environmental assessments, explore adoption dynamics across socio-economic contexts, and evaluate system-level needs for storage and grid stability using updated climate datasets.

- Climate data and models: Used CMIP5 GCMs with ISIMIP2b bias adjustment; CMIP6 ‘hot model’ issues noted. Temperature biases and model uncertainties remain. - PV performance drivers: Wind speed excluded due to high uncertainty, potentially omitting cooling and soiling dynamics. Loss factors applied as typical values rather than site-specific. - Siting constraints: Excluded detailed land suitability, environmental constraints, and micro-siting considerations; results represent technical potential over land, not fully constrained feasible potential. - Technology parameters: Used mean efficiencies per technology; real-world performance varies with climate, system design, and degradation. - Demand projections: IAM and downscaled projections are uncertain; alternative energy system pathways (e.g., 75 TW PV by 2050) yield different land needs. - System integration: Storage, transmission, grid stability, and cost dynamics were discussed qualitatively but not modeled. - Environmental and social impacts: Biodiversity, ecosystem services, and socio-economic adoption factors are only partially considered and warrant further analysis.