LIVE NEWS
  • Ledger, BBC Among 178 Organizations Hit by Crypto Malware
  • US basketball player arrested in Nigeria over 75 pounds of cannabis
  • CMS proposes more aggressive audits of Medicare Advantage| STAT
  • Microsoft says months-old bug caused global Microsoft 365 outage
  • James Bond live updates: Robert Pattinson named as new 007
  • Wildfire Lawsuits Against Utilities Fuel a Debate Over Who Should Pay
  • Tiny robots can tumble through blood vessels to help dissolve blood clots, early test shows
  • The Verge’s guide to the best gifts under $100
Prime Reports
  • Home
  • Popular Now
  • Crypto
  • Cybersecurity
  • Economy
  • Geopolitics
  • Global Markets
  • Politics
  • See More
    • Artificial Intelligence
    • Climate Risks
    • Defense
    • Healthcare Innovation
    • Science
    • Technology
    • World
Prime Reports
  • Home
  • Popular Now
  • Crypto
  • Cybersecurity
  • Economy
  • Geopolitics
  • Global Markets
  • Politics
  • Artificial Intelligence
  • Climate Risks
  • Defense
  • Healthcare Innovation
  • Science
  • Technology
  • World
Home»World»Impact of climate change on the photovoltaic power potential in Sweden across 21st century
World

Impact of climate change on the photovoltaic power potential in Sweden across 21st century

primereportsBy primereportsOctober 12, 2026No Comments14 Mins Read
Share Facebook Twitter Pinterest LinkedIn Tumblr Reddit Telegram Email
Impact of climate change on the photovoltaic power potential in Sweden across 21st century
Share
Facebook Twitter LinkedIn Pinterest Email


In this study, we evaluate PV generation in Stockholm, Gothenburg, Malmo and Kiruna over 1951–2100, and the details of the locations are presented in Supplementary note 1. Years 1951–2014 are treated as the historical reference period, and 2015–2100 as the future projection period. The analysis considers SSP1-2.6 and SSP3-7.0, two scenarios, which represent a lower-forcing sustainability pathway and a higher-forcing pathway with more limited climate change mitigation, respectively. The study also includes three different climate simulations representing different possibilities. The results are compared with the PV systems that are installed with the most common tilt angle of 15⁰ in the current Swedish PV market.

Future climate projections

Figure 1a–l summarizes the historical and projected changes in annual GHI, average air temperature, and cumulative snowfall in the four selected cities over 1951–2100. Significantly differ from the other three locations, Kiruna has the lowest GHI, an approximate 0 °C average air temperature, and almost double the snowfall. Malmo has the warmest conditions and highest GHI, together with the least snowfall, whereas Stockholm and Gothenburg occupy intermediate ranges. The projections indicate that the magnitude of change strongly depends on the emission pathways. Under SSP370, annual GHI declines by around 30–40 W/m² in all four cities, while SSP126 produces only minor changes. Average air temperature in the three southern cities rises from approximately 4–9 °C historically to 6–13 °C by 2100. In Kiruna, the corresponding range shifts from about −2.7 to −2.3 °C to approximately −1.2 to 1.6 °C. Snowfall varies considerably from year to year, but with a significant decline trend, especially under SSP370.

Fig. 1: The time series of GHI, air temperature, snowfall and annual PV generation for the four cities from 1951 to 2100.
Fig. 1: The time series of GHI, air temperature, snowfall and annual PV generation for the four cities from 1951 to 2100.

a–d Annual GHI; e–h mean air temperature; i–l total snowfall and m–p annual PV generation. Each row (each parameter) from left to right is Malmo, Gothenburg, Stockholm and Kiruna. Different colors represent different scenarios: green for SSP126, and orange for SSP370 (the historical period is the same for both scenarios). The shaded areas represent the max/min range and brown color is due to overlapping of two shades. The solid lines stand for the trendlines during the future period for each scenario. The vertical black dot line separates the historical and future periods. The installation angles are based on calculated optimal angles so the angles for the four cities are different.

All four cities are showing signs of warming, while snowfall has also decreased significantly. For Malmo, Gothenburg, and Stockholm, the temperature increase trends are comparable: approximately 0.08–0.10 °C per decade under SSP126 and 0.33–0.38 °C per decade under SSP370. The air temperature increase in Kiruna is faster, with a rate of about 0.14 and 0.51 °C per decade, respectively. Based on the fitted trend lines, temperature increases of 0.68–0.85 °C under SSP126 and 2.87–3.34 °C under SSP370 in southern Sweden (e.g., Malmö) by 2100. These values are slightly below the IPCC SSP126 estimation (1.3–2.4 °C27). For northern Sweden, represented by Kiruna, the air temperature increase reaches approximately 1.20 °C under SSP126 and 4.42 °C under SSP370 by the end of the century.

Regarding future snowfall, the selected cities show different trends. Stockholm shows the largest reduction in snowfall, around 27 mm under SSP126 and 78 mm under SSP370. The SSP126 reductions in Gothenburg and Malmo are approximately half of those in Stockholm. Under SSP370, Kiruna exhibits the smallest decrease. By 2100, snowfall in southern Sweden is expected to fall to around 2/3 of historical amounts under SSP126 and to a nearly negligible level under SSP370. Kiruna retains about 90% of its historical snowfall, corresponding to approximately 300 mm annually.

The projected evolution trend of GHI also varies under different scenarios. SSP370 produces a downward trend everywhere, strongest in Kiruna at about −0.75% per decade. The other three cities show similar declines of approximately −0.40% per decade. Under SSP126, GHI is effectively stable: Stockholm and Gothenburg show marginal increases, whereas Malmo and Kiruna decline slightly. Across all four cities, the total SSP126 change remains below 15 kW/m², equivalent to less than 0.15% per decade.

Future PV generation

Figure 1m–p shows the calculated maximum annual PV generation for each city over the 150-year period. Malmo, Gothenburg, and Stockholm occupy a similar range, generally following the order Malmo > Gothenburg > Stockholm, with yields of approximately 600–1100 kWh/kWp. The PV generation in Kiruna is considerably less, at about 350–550 kWh/kWp, or roughly half of that in the southern cities. Under SSP126, all four cities exhibit a modest increase in PV generation, with rates of 1.52–2.60 kWh/kWp per decade, corresponding to less than 0.50% per decade. The SSP370 projections are more diverse. PV generation increases in Stockholm and Kiruna, with rates of 1.30 and 3.41 kWh/kWp per decade, respectively, whereas it declines in Gothenburg and Malmö by 10.61 and 12.70 kWh/kWp by 2100. Thus, under the higher-emission pathway, the response changes from a decreasing tendency in the south to an increasing tendency toward the north. In addition, compared to the historical period, the results also show that annual PV generation is expected to become more stable.

It is noted that the PV response does not always follow the evolution of GHI. For example, GHI may decrease while PV generation increases. This apparent inconsistency reflects the competing effects of the meteorological variables. Higher air temperatures can lower module efficiency, but they may also reduce snow-related losses by promoting snowmelt and snow sliding during winter and spring. Snow coverage is an important factor in Sweden, with annual PV losses attributed to snow estimated at approximately 15%. Consequently, reduced snow losses can outweigh the adverse temperature effect and compensate for a decline in solar radiation. The net response depends on the relative strength of these competing mechanisms, motivating a separate assessment of the individual meteorological controls on PV generation.

Relationship between meteorological variables and PV generation

The relationship between meteorological variables and PV generation was evaluated using Pearson correlation coefficients, and statistical significance was assessed using two-sided tests28,29. The Pearson’s correlation coefficient and P-value were calculated using daily average values. Figure 2 summarizes the results by climate simulation, emissions scenario, and season. Seasons are defined as follows: spring (MAM, March–May), summer (JJA, June–August), autumn (SON, September–November), and winter (DJF, December–February). The variation in correlation coefficients across different simulations could be driven by differences in their driven GCMs, which differ in underlying model configurations, structural physics, and internal variability. For example, the different schemes (e.g., convective parameterizations, land-surface schemes, etc.) in the parent GCMs impact how each model simulates localized feedback mechanisms. Furthermore, each model possesses a unique trajectory of internal climate variability and differing climate sensitivities, which naturally manifest as differences in the absolute correlation values. Utilizing a multi-simulation approach allows us to quantify the uncertainties. In general, despite small value differences, the results of these simulations are similar, resulting in a consistent conclusion. PV generation is positively correlated with GHI and air temperature but negatively with snowfall. GHI is the dominant factor, with air temperature having a greater impact than snowfall. The four cities show similar patterns.

Fig. 2: Correlation coefficients between PV generation and meteorological variables across different cities.
Fig. 2: Correlation coefficients between PV generation and meteorological variables across different cities.

a Malmo; b Gothenburg; c Stockholm and d Kiruna. Associations were evaluated using Pearson correlation coefficients, and statistical significance was assessed using two-sided tests. The heatmap represents Pearson’s correlation coefficients, with darker colors indicating values closer to ±1. Star marker represents statistical significance: two stars indicate P<0.01, while one star corresponds to 0.01≤P<0.05. In summer, the row for snowfall might be blank, indicating no snowfall during that period. Exact p-value across statistical significance tests is shown in Supplementary Information Fig. 10.

A strong correlation between PV generation and GHI is evident in spring, summer and autumn, with values close to 1. Although GHI remains the most important predictor in winter, its correlation is weaker during DJF. The impact of air temperature is seasonally dependent. It is usually positive, but can become slightly negative in winter, when colder conditions prolong snow retention on the modules and suppress generation. In summer, temperature contributes little additional explanatory power and is generally not statistically significant. Weak negative temperature correlations are observed for Gothenburg and Malmo during winter in some future simulations, which may reflect changing snow conditions. Positive effects associated with snow melting require simultaneous satisfaction of two conditions: the absence of snowfall and an air temperature exceeding 0 °C. If the temperature remains below 0 °C, even an increase in temperature will not generate positive effects, as the snow melting process cannot be initiated under sub-freezing conditions. Moreover, for these two cities, which are in the southern part, there would be less snow. PV generation loss due to snow cover becomes much smaller, resembling normal operating conditions where the impact of air temperature is dominated by the negative temperature coefficient. Therefore, the positive effects of air temperature on accelerating snow melting will become smaller in the future. Then, the effect overall could vary depending on snowfall. In fact, the values are less than 0.1, which means the effect is limited.

Snowfall consistently reduces PV generation, with the strongest impact in spring. In Stockholm, Gothenburg and Malmo, snowfall primarily influences PV generation in spring, autumn and winter. In Kiruna, where the snow season is longer, snowfall can also influence PV generation in summer. In winter, snowfall effects persist longer due to lower air temperature, but statistical evidence is weaker. There are several blocks in the snow seasons without stars, but this does not represent the importance of the effects. The reason for this could be: (1) not enough samples (e.g., SN-JJA in Stockholm); (2) the variables result in co-effects. For example, in the snow season, when snow depth is more than 10 cm, there is no PV generation due to snow cover, no matter the air temperature.

According to the results above, estimated PV generation may follow a different trend from solar radiation, which is dominant when parameters are analyzed individually. Since air temperature influences snow effects, a sensitivity study is performed to assess the combined impacts of solar radiation (G) and snowfall with air temperature (SnowT). These two factors are analyzed under four conditions for future scenarios. In addition to normal simulations, (1) SSP126 (G126+SnowT126), (2) SSP370 (G370+SnowT370), (3) G126 with SnowT from SSP370 (G126+SnowT370) and (4) G370 with snowT from SSP126 (G370+SnowT126) are synthetic scenarios for sensitivity analysis only and might not represent the real future climate scenarios. This analysis was done to show that the co-effects of multiple meteorological variables could lead to different results. The annual PV generation for these conditions during 2015–2100 is presented in Fig. 3. By comparing the four scenarios, the dominant factor is determined by the importance of the future trend. If changing one factor would result in a change in trendline slope, this factor is regarded as more influential.

Fig. 3: Comparison of meteorological impacts on annual PV generation across different climate simulations in four cities.
Fig. 3: Comparison of meteorological impacts on annual PV generation across different climate simulations in four cities.

a–c Malmo, d–f Gothenburg, g–i Stockholm, and j–l Kiruna. The fluctuating lines with lighter color are the annual values under different scenarios and the solid lines with darker color are the trendlines.

Figure 2 identifies solar radiation as the primary driver of PV generation across the cities and climate simulations. These ranking changes when SnowT is included as a predictor: solar radiation remains dominant for Stockholm, Gothenburg, and Malmo in HCLIM-CNRM-ESM2-1, and for Malmo in HCLIM-MPI-ESM1-2-HR (Fig. 3a, c, d, g). In these cases, the combined impacts of air temperature and snowfall may therefore exceed those of radiation alone on PV generation. For the remaining cases (Fig. 3b, e, h, j, l), the contributions of G and SnowT are of comparable magnitude. The slope differences of the four fitted trend lines further demonstrate that changes in these drivers can substantially alter PV output. In general, the SnowT370/G126 combination is associated with increasing generation, whereas SnowT126/G370 produces a declining tendency.

In some cases, the co-effects of snow and air temperature dominate over solar radiation, as shown in Fig. 3f, i, k. The snow loss for PV generation mainly depends on the spring season (March to May), when the solar radiation is potentially good for PV generation. Even in the south of Sweden, the snow season could be until April. With global warming, there would be less snow in the spring, and solar radiation could be sufficiently utilized. The three climate simulations show significantly different attribution patterns for Stockholm and Gothenburg. Malmo, however, shows greater consistency between the two simulations: solar radiation is the leading determinant. In Kiruna, the combined effects of snowfall and air temperature are similarly important as solar radiation. This variability indicates that no individual meteorological driver dominates universally across all locations and climate simulations. Consequently, assessments based on multiple climate simulations are necessary to obtain a more reliable characterization of future PV-generation changes.

Discussions

Climate change is expected to raise air temperatures and generally reduce snowfall, although a strengthened hydrological cycle may also increase the occurrence of intense precipitation events30. Therefore, this may result in decreased solar radiation and increased cloudy days. Consistent with this possibility, annual GHI declines under SSP370 in all four cities. In contrast, under the lower-forcing SSP126 pathway, the response is spatially heterogeneous: Stockholm and Gothenburg show modest increases, whereas Malmo and Kiruna exhibit decreases.

In Sweden, future PV output change cannot be solely predicted based on GHI due to its long snow season. Although PV generation is strongly and positively associated with solar radiation, air temperature and snow-related processes exert competing influences. Elevated air temperature can lower conversion efficiency due to the temperature effect on the PV cell, but it can also promote snowmelt and snow sliding, thereby reducing snow-induced losses. The relative contribution of these mechanisms differs among locations and climate simulations. Solar radiation is generally more influential in southern Sweden, while snowfall and air temperature become increasingly important toward the north. In some regions of Sweden, the combined effect of snowfall and air temperature is comparable to that of solar radiation. Therefore, without considering snow processes could lead to biased estimation of future PV performance.

Figure 4 summarizes PV generation, optimal installation angle (for maximizing PV generation) and the corresponding gain relative to the commonly used commercial angle and the P50 values of seasonal generation for each city and scenario. Across the study area, future annual PV generation is approximately 5% higher than that of the historical period, equivalent to an increase of about 33  kWh/kWp. This result differs from studies that projected future reductions, likely since those analyzes did not represent snow-related losses15,16,17,18. A comparison between radiation treatments of the previous methods and this study is provided in Supplementary note 2. PV generation under SSP370 is generally higher than under SSP126, especially in Kiruna. Although summer supplies more than 40% of the annual PV yield in southern Sweden, the projected increase is mainly contributed by the snow season, especially spring, as indicated by the rounded markers. The future optimal tilt angle increases slightly in the south but decreases toward northern Sweden, where year-to-year variation is also greater. Future annual PV output is generally less variable than in the historical period. The median projected optimum tilt is approximately 35° for all cities. Using the city-specific optimal tilt angle rather than the commercial angle typically raises annual yield by 3–6%, or about 20–40 kWh/kWp.

Fig. 4: Comparison of key performance indicators of future PV systems under different scenarios in four cities.
Fig. 4: Comparison of key performance indicators of future PV systems under different scenarios in four cities.

a Optimized solar installation angle; b PV generation; c increased power generation compared to commercial angle; d the median (P50) of seasonal PV generation. Four different types of markers represent four seasons (spring: March-April-May, summer: June-July-August, autumn: September-October-November and winter: December-January-February), and different colors mean different scenarios.

Extreme weather conditions, particularly during the snow season, can produce extended periods of zero PV output. These periods are important indicators of system adequacy because they influence grid planning and may require additional energy storage capacity, thereby increasing the cost of reliable PV integration. The total number of days without PV generation and the maximum consecutive days without PV generation are shown in Fig. 5. Relative to the historical baseline, both future scenarios generally reduce the frequency and duration of zero-output periods. Especially under SSP370, these values in Stockholm, Gothenburg, and Malmo could be half of their historical levels. This overall decline suggests lower variability in future PV production and potentially reduced requirements for backup storage. However, prolonged shortages without PV generation lasting approximately 1 to 2 consecutive weeks remain possible in some years. The projections also contain numerous outliers under both scenarios, indicating that individual extreme years may still produce unusually high intermittency. For example, some SSP370 values for Stockholm exceed the historical range. Thus, although the average conditions become more favorable, occasional climate extremes could create more demanding operating conditions for PV systems.

Fig. 5: Zero PV generation days within a year.
Fig. 5: Zero PV generation days within a year.

a The total number of days; b max number of consecutive days. For Kiruna, the polar nights (22 days) are included.

Share. Facebook Twitter Pinterest LinkedIn Tumblr Email
Previous ArticlePLTR Price Prediction: Goldman and Barclays Turn Bullish as Technicals Flash Extreme Caution
Next Article AHA sues to halt 340B rebate pilot from Trump administration
primereports
  • Website

Related Posts

World

Wildfire Lawsuits Against Utilities Fuel a Debate Over Who Should Pay

November 16, 2026
World

Analysis: Climate-finance pledges fall 70% as countries meet at COP31 

November 16, 2026
World

Behind the COP31 fanfare, Türkiye is worried about water

November 16, 2026
Add A Comment
Leave A Reply Cancel Reply

Top Posts

Threat of further violence looms after Mexican cartel rampage

February 25, 2026116 Views

‘Two-sided risk’ Medicare Advantage plans improve patient outcomes

February 24, 202673 Views

Police chief urges those ‘who know nothing about Northern Ireland’ not to stir up disorder via social media – UK politics live | Politics

June 9, 202631 Views
Stay In Touch
  • Facebook
  • YouTube
  • TikTok
  • WhatsApp
  • Twitter
  • Instagram
Latest Reviews

Subscribe to Updates

Get the latest tech news from FooBar about tech, design and biz.

PrimeReports.org
Independent global news, analysis & insights.

PrimeReports.org brings you in-depth coverage of geopolitics, markets, technology and risk – with context that helps you understand what really matters.

Editorially independent · Opinions are those of the authors and not investment advice.
Facebook X (Twitter) LinkedIn YouTube
Key Sections
  • World
  • Crypto
  • Cybersecurity
  • Geopolitics
  • Global Markets
  • Artificial Intelligence
All Categories
  • Artificial Intelligence
  • Climate Risks
  • Crypto
  • Cybersecurity
  • Defense
  • Economy
  • Geopolitics
  • Global Markets
  • Healthcare Innovation
  • Politics
  • Popular Now
  • Science
  • Technology
  • World
  • About Us
  • Contact Us
  • Privacy Policy
  • Terms & Conditions
  • Disclaimer
  • Cookie Policy
  • DMCA / Copyright Notice
  • Editorial Policy

Sign up for Prime Reports Briefing – essential stories and analysis in your inbox.

By subscribing you agree to our Privacy Policy. You can opt out anytime.
Latest Stories
  • Ledger, BBC Among 178 Organizations Hit by Crypto Malware
  • US basketball player arrested in Nigeria over 75 pounds of cannabis
  • CMS proposes more aggressive audits of Medicare Advantage| STAT
© 2026 PrimeReports.org. All rights reserved.
Privacy Terms Contact

Type above and press Enter to search. Press Esc to cancel.