Abstract Photovoltaic (PV) systems integrated with green roofs have attracted increasing research interest due to their potential influence on rooftop microclimatic conditions and photovoltaic operating performance. This study experimentally investigated the thermal and electrical behavior of two identical PV modules installed above green and asphalt roof surfaces under real operating conditions in a Central European climate. Rear-side module temperatures and meteorological parameters were monitored, while electrical performance was evaluated using on-site I–V curve measurements. The observed rear-side temperature differences ranged from 0.01 °C to 0.86 °C during the monitored short-term summer periods. A representative I–V measurement indicated approximately 13% higher instantaneous maximum power output for the PV module installed above the green roof configuration under comparable operating conditions. However, the electrical results should be interpreted cautiously due to short-term environmental variability and irradiance-related uncertainty during consecutive field measurements. The presented results correspond to a short-term summer field-monitoring study and should not be generalized to annual photovoltaic performance without extended long-term multi-season experimental validation. The scientific contribution of this study lies in the synchronized side-by-side evaluation of identical PV modules using combined rear-side thermal monitoring and in-situ electrical characterization under real operating conditions. 1. Introduction Rooftop photovoltaic (PV) systems are increasingly deployed as a practical pathway to decarbonize buildings and urban energy systems. Their integration into existing roof infrastructure is attractive due to the available roof surface area and proximity to electricity demand [ 1]. However, the implementation of distributed energy systems in urban environments is affected by spatial, technical, and operational constraints [ 2]. Real-world PV performance is also influenced by environmental stressors such as soiling, thermal loading, and seasonal variability [ 3]. Among these factors, module operating temperature is particularly important because it directly affects voltage and electrical output [ 4]. Broader environmental conditions, including solar irradiance, ambient temperature, atmospheric effects, and extreme weather events, further influence PV productivity under real operating conditions [ 5]. Temperature-related efficiency losses have been confirmed experimentally for both monocrystalline and polycrystalline photovoltaic panels [ 6]. Active cooling strategies, including water-based cooling systems, have therefore been investigated as possible methods for improving PV module performance under high-temperature conditions [ 7]. In the built environment, the thermal behavior of PV modules is additionally affected by roof geometry, surrounding surfaces, ventilation conditions, and urban microclimate interactions [ 8]. In parallel, green roofs have become a widely adopted nature-based solution supporting climate adaptation in cities. They can moderate roof surface temperatures, retain moisture, and enhance evapotranspiration, thereby influencing the thermal behavior of buildings and roof envelopes [ 9]. Previous reviews have shown that green roof performance depends strongly on substrate depth, vegetation type, water availability, and structural configuration [ 10]. Experimental studies conducted in Central European conditions have also demonstrated that the thermal regime of extensive green roofs varies significantly with seasonality [ 11]. Evapotranspiration and vegetation characteristics play an important role in evaporative cooling, particularly under warm weather conditions [ 12]. Numerical and experimental studies further indicate that green roof thermal behavior is governed by coupled heat and moisture transfer processes [ 13]. Rainwater retention and substrate moisture conditions can also affect membrane temperatures and evapotranspiration intensity [ 14]. The combination of PV systems with vegetated roofs, commonly referred to as biosolar roofs, has emerged as a promising integrated building strategy. A systematic review by Talwar et al. showed that PV–green roof performance depends strongly on climatic conditions, PV mounting height, vegetation characteristics, substrate moisture, and the PV-to-green-roof ratio [ 15]. GIS-based assessments have further highlighted the potential of combining photovoltaic and green roof systems within urban roof areas [ 16]. Simulation studies suggest that green roofs and PV systems may jointly contribute to building energy savings and urban climate mitigation when properly integrated [ 17]. However, field observations show that the thermal response of PV roofs can vary significantly during heatwave conditions and is not always linearly related to roof surface cooling [ 18]. Recent review work on photovoltaic–thermal systems combined with green roofs also emphasizes that their benefits depend on system design, climate, and local operating conditions [ 19]. Recent experimental evidence confirms that biosolar roof effects are highly site-specific. Möslinger et al. reported average daytime temperature reductions beneath a green roof ranging from 0.5 °C to 2.2 °C depending on season, with peak differences up to 8 °C during hot summer days in a two-year field study in Austria [ 20]. These results indicate that the magnitude of PV–green roof cooling depends not only on vegetation and substrate characteristics, but also on module geometry, airflow conditions, and seasonal moisture availability. Therefore, direct comparison between studies requires careful consideration of the measured parameter, such as roof surface temperature, air temperature beneath the module, front-side module temperature, or rear-side module temperature. A critical methodological issue is the way in which thermal measurements are linked to electrical performance. Several studies have investigated the influence of rooftop PV systems on urban thermal microclimates and roof surface temperatures [ 21]. Other experimental work has focused on methods for controlling rooftop PV surface temperature through cooling or ventilation strategies [ 22]. However, electrical performance evaluation under real operating conditions also requires careful interpretation of I–V curve measurements [ 23]. PV module temperature is affected by incident solar spectrum and time-dependent irradiance conditions, which can influence the measured electrical response [ 24]. In addition, I–V curve interpretation can be affected by faults, mismatch, and transient operating conditions [ 25]. Long-term PV yield studies further show that annual performance depends on seasonal weather variability and local climate [ 26]. High-resolution simulations also confirm that PV generation profiles must be evaluated together with temporal load and irradiance patterns when assessing rooftop PV potential [ 27]. Despite the growing body of research, there remains limited experimental evidence for biosolar roof behavior under Central European climatic conditions using controlled side-by-side configurations. Many previous studies focus primarily on roof surface temperatures, modeled scenarios, or general rooftop PV potential rather than synchronized rear-side PV module temperature measurements [ 1, 18]. Fewer studies combine rear-side thermal monitoring with on-site I–V curve characterization under real operating conditions [ 23, 27]. This methodological gap is important because rear-side module temperatures may show smaller differences than roof surface temperatures, especially under naturally ventilated rooftop configurations. The novelty of this study lies in the synchronized side-by-side experimental evaluation of identical photovoltaic modules installed above green and asphalt roof surfaces under real operating conditions in a Central European climate. Unlike many previous studies focused primarily on roof-level or surface temperature analysis, the present investigation combines synchronized rear-side thermal monitoring with in-situ I–V curve characterization using identical PV modules exposed to comparable environmental conditions. The study therefore provides experimentally controlled field-based insight into the short-term microclimatic interaction between vegetated roof systems and rooftop photovoltaic performance. 2.1. Experimental Site and Roof Configurations The experimental investigation was conducted on the roof of a university building located in Košice, Slovakia, representing typical Central European climatic conditions. The experimental setup was designed to enable a direct comparison between photovoltaic (PV) modules installed above two different roof surface types: a conventional asphalt roof and an extensive green roof system ( Figure 1). The green roof consisted of a vegetated substrate layer with low-growing vegetation, providing moisture retention and evapotranspiration potential, while the reference roof surface was a standard asphalt waterproofing layer. Both roof surfaces were exposed to identical ambient conditions, including solar radiation, wind, and air temperature. The vertical distance between the rear side of the PV modules and the roof surfaces ranged from approximately 100 mm at the lower edge to 465 mm at the upper edge, as illustrated in Figure 2. The same mounting geometry was used for both roof configurations to ensure comparable natural airflow and convective cooling conditions beneath the PV modules. The selected installation spacing is consistent with commonly applied small-scale rooftop biosolar PV mounting practices reported in experimental studies. 2.2. Photovoltaic Modules The experimental setup employed two identical polycrystalline photovoltaic modules with a nominal power output of 260 Wp. Using identical PV modules ensured that any observed differences in thermal or electrical performance could be attributed to the underlying roof surface rather than to module-specific characteristics. Both modules were installed using the same mounting structure, orientation, and tilt angle to eliminate geometric effects on solar irradiance. The electrical characteristics of the modules were periodically verified through I–V curve measurements to confirm consistent performance behavior throughout the monitoring period. 2.3. Temperature Measurement System and Sensor Placement Rear-side PV module temperatures were measured using a multi-channel data acquisition system ALMEMO ପ୍ପ 500 ( Figure 3), which served as the central unit for temperature data collection. A total of eight contact temperature sensors (type ZA9020-FS THERMO R2E4) were deployed within the experimental setup. For each PV module, two temperature sensors were installed on the rear side: One positioned at the upper section of the module; One positioned at the lower section of the module. The manufacturer-specified accuracy of the ZA9020-FS THERMO R2E4 ( Figure 4) temperature sensors is ±0.2 °C within the relevant operating range. To minimize random measurement fluctuations, two sensors were installed on the rear side of each PV module and their average value was used to represent module operating temperature. This averaging approach reduces the influence of localized thermal heterogeneity and improves the robustness of the measured temperature values. Calibration consistency between sensors was verified prior to installation to ensure comparable measurement behavior across both PV modules. It should be noted that the measured rear-side temperatures do not directly represent the actual photovoltaic cell temperature. Under real operating conditions, the internal PV cell temperature is typically several degrees Celsius higher than the measured rear-side module temperature due to internal heat generation and thermal resistance within the module layers. Nevertheless, rear-side temperature monitoring provides a practical and consistent experimental indicator for comparative evaluation of PV thermal behavior under synchronized operating conditions. The sensors were fixed using thermally conductive adhesive tape to ensure good thermal contact while minimizing external interference. The average value of the two rear-side sensors was used to represent the operating temperature of each PV module. Additional temperature sensors were installed to characterize the thermal behavior of the roof surfaces: Beneath the PV modules in shaded conditions; In front of the modules in sun-exposed areas, for both the green roof and the asphalt roof. This sensor ( Figure 5) arrangement allowed for the assessment of surface temperature differences and their influence on the microclimatic conditions beneath the PV installations. 2.4. Meteorological Data Acquisition Meteorological data were obtained from a weather station located in close proximity to the experimental site. The recorded parameters included: ambient air temperature; global solar radiation; relative humidity; average wind speed; maximum wind speed; precipitation. All meteorological parameters were recorded at five-minute intervals, synchronized with the temperature measurements acquired by the ALMEMO ପ୍ପ 500 system. This synchronization enabled a consistent dataset for evaluating the influence of environmental conditions on PV module operating temperatures. 2.6. Electrical Performance Measurements The electrical performance of the PV modules was assessed using an MP-11 ( Figure 8 I–V curve tracer (EKO Instruments Co., Ltd., Tokyo, Japan), which enabled the measurement of current–voltage characteristics directly on-site. I–V curves were recorded during defined measurement windows for each monitoring day, typically at five-minute intervals, under comparable irradiance conditions. From the measured I–V curves, key electrical parameters such as open-circuit voltage, short-circuit current, and maximum power output were derived. These measurements were used to evaluate the influence of operating temperature on PV module performance. 2.7. Data Processing and Analysis Temperature and meteorological data were processed using a structured data analysis workflow. For each monitoring day, a specific measurement window was defined based on the on-site experimental schedule, representing periods with stable operating conditions. Within each window, average rear-side PV module temperatures were calculated for both the green and asphalt roof installations. The temperature difference (ΔT) between the two configurations was defined as the difference between the average rear-side temperatures of the asphalt roof and green roof installations. This approach enabled a direct and consistent comparison of the thermal behavior of PV modules installed above different roof surface types. Basic descriptive statistical analysis was performed for each defined measurement window to evaluate the robustness of the observed temperature differences. For each monitoring day, the mean rear-side temperature and the corresponding standard deviation were calculated for both configurations. Although the absolute magnitude of the observed temperature differences was relatively modest, the green roof configuration consistently exhibited lower average rear-side temperatures during most monitoring periods. Nevertheless, temperature differences approaching the manufacturer-specified sensor accuracy range (±0.2 °C) should be interpreted cautiously. This supports the reliability of the detected cooling tendency under the investigated operating conditions. To further evaluate measurement reliability and repeatability, the paired rear-side temperature sensors installed on each PV module were analyzed throughout the monitoring campaign. For each roof configuration, the average value of the upper and lower rear-side sensors was used to represent the module operating temperature, thereby reducing the influence of localized thermal heterogeneity and short-term sensor fluctuations. Additional repeatability analysis was performed using the mean absolute error (MAE) and root mean square error (RMSE) calculated between paired sensors installed on the same PV module. The resulting repeatability indicators are summarized in Table 2. The calculated MAE and RMSE values primarily reflect the natural thermal non-uniformity between the upper and lower rear-side regions of the PV modules caused by convective airflow and localized heating effects, rather than pure sensor measurement uncertainty alone. The present analysis was primarily intended to evaluate short-term repeatability and consistency of the synchronized temperature measurements under real operating conditions rather than to establish high-confidence statistical significance for small absolute temperature differences. Therefore, the reported ΔT values should be interpreted as indicative short-term thermal trends within the context of the measurement uncertainty and experimental limitations. Considering the relatively small magnitude of certain observed ΔT values, temperature differences close to the manufacturer-specified sensor accuracy range (±0.2 °C) were interpreted with caution. Therefore, the presented results should primarily be understood as evidence of a consistent cooling tendency observed under synchronized real operating conditions rather than as an exact quantification of absolute thermal gain. Data visualization and statistical analyses were performed using spreadsheet-based tools. Scatter plots were used to investigate potential relationships between temperature differences and selected environmental parameters, including wind speed and global solar radiation. 3.1. Rear-Side PV Module Temperature Difference Figure 9 presents the rear-side photovoltaic module temperature difference (ΔT) between installations above asphalt and green roof surfaces for all monitored days. For each day, the temperature difference was calculated as the average value within the corresponding measurement window, which varied depending on the on-site measurement schedule but consistently represented periods with relatively stable operating conditions. Across the monitored days, the PV modules installed above the green roof consistently exhibited lower rear-side temperatures compared to those installed above the asphalt roof. The observed temperature differences ranged from approximately 0.01 °C up to 0.86 °C. Although the magnitude of the temperature reduction varied between individual days, a cooling tendency associated with the green roof configuration was observed in all monitored periods. Descriptive statistical comparison indicates that, in most monitoring windows, the observed ΔT values exceeded the expected short-term measurement variability, suggesting a consistent cooling tendency under the investigated operating conditions. In certain cases, however, the temperature differences were comparable to the manufacturer-specified sensor accuracy range (±0.2 °C) and should therefore be interpreted with appropriate caution. In order to provide context for the measured temperature differences, the main environmental parameters recorded by the nearby meteorological station during each measurement window are summarized in Table 3. These parameters include ambient air temperature, wind speed, and global solar radiation, which are known to influence PV module operating conditions. 3.2. Influence of Environmental Conditions on Temperature Differences The environmental parameters corresponding to each measurement window are summarized in Table 3, allowing for a more comprehensive interpretation of the relationships presented in Figure 10. To further investigate the factors influencing the observed temperature differences, the relationship between the rear-side PV module temperature difference and selected environmental parameters was analyzed. Figure 9 illustrates the relationship between the average wind speed and the rear-side temperature difference calculated for each measurement window. No strong linear correlation between wind speed and temperature difference was identified across the monitored days. The results indicate that no single environmental parameter dominates the observed temperature difference. A similar behavior was observed when analyzing the relationship between temperature difference and global solar radiation. Although higher radiation levels increased the absolute operating temperature of the PV modules, the relative temperature difference between the green and asphalt roof installations remained limited. These results suggest that the cooling potential of green roofs beneath photovoltaic modules is influenced by complex heat transfer mechanisms rather than by a single dominant environmental parameter. 3.3. Electrical Performance of PV Modules In addition to the thermal analysis, the electrical performance of the photovoltaic modules installed above the green and asphalt roof surfaces was evaluated using on-site I–V curve measurements. The objective of this analysis was to assess whether the observed differences in operating temperature were reflected in the electrical behavior of the PV modules under real operating conditions. I–V characteristics were measured using a portable MP-11 I–V curve tracer during the defined measurement windows for each monitoring day. Measurements for both PV installations were conducted within short time intervals under nearly identical irradiance and ambient conditions to ensure direct comparability. During the representative measurement shown in Figure 11, the measured global irradiance remained stable and differed by less than 2% between the two consecutive measurements. No translation of the measured I–V curves to standard test conditions was applied, as the objective of this study was to compare the real operating performance of both installations under identical environmental exposure rather than standardized laboratory conditions. From the recorded I–V curves, key electrical parameters, including open-circuit voltage (Voc), short-circuit current (Isc), and maximum power output (Pmax), were derived. Figure 11 presents representative I–V characteristics of the photovoltaic modules installed above the green and asphalt roof surfaces under comparable operating conditions. The presented data correspond to a representative measurement conducted on 6 August 2025 at 13:47 and are intended primarily as illustrative field observations under synchronized operating conditions. Nevertheless, the presented electrical measurements represent discrete short-term observations rather than long-term statistically averaged photovoltaic performance indicators. This difference may be partially associated with the slightly higher open-circuit voltage (Voc) observed for the green roof installation. However, short-term environmental variability and minor irradiance differences between consecutive measurements may also have contributed to the observed electrical response. As shown in Figure 11, the shape of the I–V curves remains similar for both installations, indicating comparable irradiance conditions during the measurements. The slight horizontal shift of the curve corresponding to the green roof configuration reflects the temperature-related voltage increase and the associated improvement in maximum power output. It should be noted that the electrical performance measurements were performed during discrete monitoring periods rather than continuous long-term monitoring. Therefore, the presented I–V characteristics represent instantaneous operating conditions. Nevertheless, the observed electrical behavior remained qualitatively consistent with the thermal observations obtained during the synchronized monitoring periods. However, the presented measurements should not be interpreted as statistically robust long-term evidence of photovoltaic efficiency enhancement. Detailed I–V characteristics recorded at five-minute intervals for all monitoring days support the trends observed in the representative I–V curves and provide a comprehensive dataset for future comparative analyses. These supplementary measurements indicate that the representative I–V curve presented in Figure 11 should be interpreted primarily as an illustrative field observation under synchronized operating conditions rather than as a statistically averaged electrical performance indicator. To facilitate a clear interpretation of the results in accordance with the thermal analysis, the Maximum Power Points (MPP) are designated as point A for the asphalt roof installation (black curve) and point B for the green roof installation (red curve). The maximum power output (Pmax) is determined by the specific point on the I–V curve where the product of voltage and current reaches its maximum value: Pmax = Vmpp · Impp (1) where: Pmax is the maximum power output of the PV module [W]; Vmpp is the voltage at the maximum power point [V]; Impp is the current at the maximum power point [A]. As shown in Figure 11 and summarized in Table 4, the module associated with the green roof (point B) demonstrated a higher power output (132.94 W) compared to the asphalt reference (point A, 117.63 W). The observed performance difference may be partially influenced by the cooling effect associated with the green roof configuration. However, the instantaneous electrical response was also affected by short-term environmental variability and potential minor irradiance inconsistencies between the consecutive measurements. For crystalline silicon modules, the efficiency is highly sensitive to operating temperature, where the open-circuit voltage (Voc) decreases as the cell temperature (Tcell) rises, following the relationship: Voc(T) = Voc,stc · [1 + β · (Tcell − Tstc)] (2) where: Voc(T) is the open-circuit voltage at the actual operating temperature [V]; Voc,stc is the open-circuit voltage under Standard Test Conditions (STC) [V]; β is the temperature coefficient of the open-circuit voltage [/°C or %/°C]; Tcell is the actual operating temperature of the PV cell [°C]; Tstc is the reference temperature under STC (typically 25 °C). Table 4. Comparative electrical and thermal parameters for the representative measurement (6 August 2025, 13:47). Table 4. Comparative electrical and thermal parameters for the representative measurement (6 August 2025, 13:47). Parameter Symbol Unit Asphalt Roof (A) Green Roof (B) Difference Maximum Power Pmax [W] 117.63 132.94 13.00% Open-Circuit Voltage Voc [V] 34.16 34.35 0.60% Short-Circuit Current Isc [A] 4.51 5.24 16.20% Fill Factor FF [-] 0.763 0.739 −3.10% Rear-side Temperature Trear [°C] 54.3 52.0 2.3 °C Note: The rear-side temperature (Trear) values were extracted from the comprehensive thermal dataset. The relatively large difference observed in short-circuit current (Isc) between the two measurements cannot be attributed solely to the measured rear-side temperature difference. Since irradiance was obtained from the nearby meteorological station rather than measured directly in the plane of each PV module, the observed Isc difference was likely influenced by short-term irradiance variability and transient environmental effects during the consecutive field measurements. During the representative I–V measurements presented in Figure 11, the nearest meteorological station record to the I–V acquisition time was obtained at 13:45 and indicated a global radiation value of 366.3 W/m 2. The average global radiation over the full monitoring window from 13:15 to 14:15 was 623.55 W/m 2. Since irradiance was recorded by a nearby meteorological station rather than directly at the plane of each PV module, short-term irradiance variability and transient environmental effects may have contributed to the observed differences in electrical response. Therefore, the presented electrical results should be interpreted primarily as descriptive field observations rather than as statistically robust quantification of temperature-induced photovoltaic performance enhancement. 4. Discussion The observed rear-side PV module temperature differences between the green and asphalt roof installations were relatively modest, with maximum values below 1 °C. However, this magnitude is consistent with the experimental configuration and the nature of the measured parameters. The temperature sensors were installed on the rear side of the PV modules, where convective heat transfer plays a dominant role, resulting in reduced temperature gradients compared to front-side or roof surface measurements. The cooling effect can be attributed to evapotranspiration and the thermal properties of the vegetated substrate. However, the present investigation was primarily designed as an experimental field-monitoring study rather than a detailed thermodynamic or energy-balance analysis. Consequently, parameters such as convective heat transfer coefficients, evapotranspiration rates, thermal resistance modeling, and detailed energy-balance calculations were not explicitly quantified. Future research integrating experimental measurements with numerical and simulation-based approaches would therefore be beneficial for improving the mechanistic understanding of biosolar roof thermal interactions. The exploratory scatter-plot analysis did not reveal a strong direct linear relationship between the observed temperature differences and individual environmental parameters such as wind speed. However, due to the limited dataset and the absence of multivariable statistical modeling, the present study cannot conclusively determine the relative contribution of coupled environmental factors influencing PV thermal behavior. Future investigations employing multivariable regression analysis, coupled environmental modeling, or machine-learning approaches may provide deeper insight into the interaction between climatic variables and biosolar roof thermal performance. Table 5 summarizes representative findings reported in selected previous biosolar roof studies together with the results obtained in the present investigation. Compared to several previously published studies, the temperature differences observed in the present study were generally smaller. Previous studies reported reductions ranging from approximately 0.5 °C to 2.2 °C under long-term or optimized biosolar roof configurations. This discrepancy may be associated with differences in climatic conditions, vegetation characteristics, PV mounting geometry, monitoring duration, and the rear-side sensor placement used in the present experimental setup. The comparison further highlights the strong site-specific dependence of biosolar roof thermal behavior and photovoltaic performance. Considering the manufacturer-specified sensor accuracy (±0.2 °C) and the averaging of two rear-side sensors per module, the reported temperature differences should be interpreted within the context of measurement uncertainty. While several daily ΔT values exceeded the sensor tolerance range, certain monitoring periods, particularly on 12 August 2025, 15 August 2025, and 20 August 2025, produced temperature differences comparable to or below the practical measurement resolution and should therefore be interpreted with caution. The particularly small temperature difference observed on 15 August 2025 (ΔT ≈ 0.008 °C) may be associated with the combined influence of relatively high ambient temperature, moderate wind conditions, and reduced microclimatic contrast between the two roof configurations during the selected measurement window. Since the measured ΔT value was substantially below the practical sensor resolution range, this monitoring period should be interpreted primarily as a near-equilibrium operating condition rather than as evidence of a measurable cooling effect. Nevertheless, the consistent tendency toward lower rear-side temperatures for the green roof configuration observed across the monitoring campaign suggests the presence of a systematic microclimatic effect rather than purely random measurement variability. Therefore, temperature differences comparable to or below the manufacturer-specified sensor accuracy range should be interpreted primarily as indicative short-term thermal trends rather than statistically robust absolute temperature reductions. Despite the relatively modest magnitude of the observed rear-side temperature reductions, the present study provides experimentally controlled field data obtained under synchronized real operating conditions. The results highlight the sensitivity of biosolar roof thermal interactions to local climatic conditions, measurement methodology, and installation geometry. In this context, the study contributes to improving the experimental understanding of short-term PV–green roof interactions under Central European summer conditions. 5. Limitations and Future Research The present study has several limitations that should be considered when interpreting the reported results. First, the monitoring campaign was conducted during a limited number of summer measurement days under relatively stable operating conditions. Although this approach ensured direct comparability between the two roof configurations, it does not capture the full seasonal variability influencing photovoltaic performance. Second, the temperature analysis was primarily based on rear-side PV module measurements. Rear-side temperatures are relevant for evaluating convective cooling effects and realistic operating conditions; however, they generally exhibit smaller temperature gradients than front-side module surfaces or roof surface temperatures. Consequently, the observed temperature differences may be lower than values reported in studies focused on surface-level thermal behavior. Furthermore, the experimental setup involved a single pair of photovoltaic modules installed in close proximity and monitored during selected summer periods. While this configuration was intentionally chosen to minimize spatial variability and ensure comparable boundary conditions, the absence of replicated PV modules, seasonal measurements, and long-term monitoring limits the broader generalizability of the presented results. Future studies would therefore benefit from larger sample sizes, extended monitoring periods, confidence interval evaluation, and uncertainty propagation analysis in order to improve the statistical robustness of the observed thermal differences. Future research should include long-term monitoring across multiple seasons and varying weather conditions. Additional experiments involving active cooling methods, such as controlled water application on PV module surfaces, may further improve understanding of the relationship between module temperature and electrical performance under transient operating conditions. Future research will also focus on integrating the experimental measurements with numerical and BIM-based simulation models. This approach may improve predictions of thermal behavior, photovoltaic energy yield, and building-scale biosolar roof performance under different design scenarios, while also enabling more detailed evaluation of convective heat transfer mechanisms, evapotranspiration effects, and energy-balance behavior within roof systems. Despite these limitations, the present study provides experimental insight into the thermal interaction between green roofs and rooftop photovoltaic systems under real operating conditions. The results contribute to the growing body of experimental biosolar roof research conducted under Central European climatic conditions. From a seasonal perspective, the benefits of green roof integration with photovoltaic systems are not uniformly distributed throughout the year. During summer periods, when solar irradiance is high and PV systems reach peak production, the cooling effect of the green roof can contribute to a measurable increase in electrical output. However, these periods are often associated with reduced on-site electricity demand in residential and certain commercial buildings, which may lead to surplus energy generation that is not fully utilized. In contrast, during winter conditions, when electricity demand is typically higher, the overall photovoltaic production is significantly reduced due to lower solar irradiance and shorter daylight hours. Under these conditions, the thermal influence of the green roof on PV module performance becomes negligible, and the potential for performance enhancement is limited. Consequently, while green roofs may contribute to moderated PV operating conditions during favorable summer periods, realistic assessment of their long-term energy and operational benefits requires extended annual monitoring and comprehensive energy-yield analysis under varying climatic and building operating conditions. This study experimentally investigated the thermal and electrical behavior of identical photovoltaic modules installed above green and conventional asphalt roof surfaces under real summer operating conditions in a Central European climate. Based on the obtained results, the following conclusions can be drawn: Rear-side temperature differences between the investigated roof configurations ranged from 0.01 °C to 0.86 °C during the monitored summer periods. The photovoltaic module installed above the green roof consistently exhibited slightly lower rear-side operating temperatures compared to the asphalt roof configuration, indicating a consistent microclimatic cooling tendency associated with the vegetated roof surface. The representative I–V measurement indicated higher instantaneous maximum power output for the PV module installed above the green roof configuration under comparable operating conditions. The representative I–V measurement indicated higher instantaneous maximum power output for the PV module installed above the green roof configuration under comparable operating conditions. However, the observed electrical difference should be interpreted cautiously, as the response was partly associated with an unexplained increase in short-circuit current together with short-term environmental variability during consecutive field measurements. The observed thermal and electrical effects were influenced by multiple coupled environmental factors, including solar radiation intensity, ambient temperature, wind conditions, roof surface properties, and local microclimatic interactions. Comparison with previous biosolar roof studies confirmed that the magnitude of the observed cooling effects strongly depends on climatic conditions, vegetation characteristics, PV mounting geometry, monitoring duration, and measurement methodology. From a practical perspective, the present results indicate only modest short-term thermal effects under the investigated summer operating conditions. Therefore, the findings should not be interpreted as direct evidence of substantial annual photovoltaic energy-yield improvement or large-scale biosolar roof optimization without further long-term experimental and simulation-based validation. The present investigation represents short-term summer monitoring under real operating conditions. Therefore, the results should not be generalized to annual photovoltaic performance without long-term and multi-season experimental validation. The scientific contribution of the present study lies primarily in the controlled synchronized field methodology combining rear-side thermal monitoring with in-situ I–V characterization of identical PV modules installed above different roof surface types under real operating conditions. Future research should focus on long-term monitoring, seasonal analysis, uncertainty evaluation, and annual energy-yield assessment in order to better quantify the practical energy benefits of biosolar roof systems. Author Contributions Conceptualization, P.K. and M.K.; Methodology, P.K. and F.V.; Software, P.K.; Validation, P.K. and F.V.; Formal analysis, F.V. and Z.V.; Investigation, P.K., F.V. and M.K.; Resources, P.K. and M.K.; Data curation, P.K. and F.V.; Writing—original draft, P.K. and F.V.; Writing—review & editing, F.V. and M.K.; Visualization, P.K.; Supervision, F.V. and Z.V.; Project administration, Z.V.; Funding acquisition, Z.V. All authors have read and agreed to the published version of the manuscript. Funding This research is a part of the project Research and Innovation for Decarbonization of Construction through Low-Emission Construction Processes and Materials in a Circular Economy; Acronym: DEKAST; Project Code: 09I04-03-V02-00051; Program: Recovery and Resilience Plan of the Slovak Republic and the VEGA research project Nr. 1/0382/25 Building Constructions with Vegetated and Wetland Layers in the Context of Sustainable Building Renovation. Data Availability Statement The original contributions presented in this study are included in the article. Further inquiries can be directed to the corresponding author. Acknowledgments During the preparation of this manuscript, the authors used an AI-based language model for text editing and language refinement. The authors have reviewed and edited the output and take full responsibility for the content of this publication. Abbreviations The following abbreviations are used in this manuscript: PV Photovoltaic I-V Current-Voltage BIM Building Information Modelling HVAC Heating, Ventilation and Air Conditioning Wp Watt-peak Voc Open-circuit voltage Isc Short-circuit current Pmax Maximum power output ΔT Temperature difference References Long, Y.; Xu, X.; Huo, Z. Urban rooftop photovoltaic potential model: A study on assessment methods and model framework. Energy Build. 2025, 345, 116138. [ Google Scholar] [ CrossRef] Marrone, P.; Fiume, F.; Laudani, A.; Montella, I.; Palermo, M.; Fulginei, F.R. Distributed Energy Systems: Constraints and Opportunities in Urban Environments. Energies 2023, 16, 2718. [ Google Scholar] [ CrossRef] Uykan, O.; Çelik, G.; Birgül, A. Quantifying the Impact of Soiling and Thermal Stress on Rooftop PV Performance: Seasonal Analysis from an Industrial Urban Region in Türkiye. Sustainability 2025, 17, 8038. [ Google Scholar] [ CrossRef] Bezaatpour, J.; Gholizadeh, T.; Bezaatpour, M.; Ebadollahi, M.; Ghaebi, H. Strategic mitigation of temperature-induced efficiency losses in large-scale photovoltaic facades. Energy 2025, 334, 137792. [ Google Scholar] [ CrossRef] Bamisile, O.; Acen, C.; Cai, D.; Huang, Q.; Staffell, I. The environmental factors affecting solar photovoltaic output. Renew. Sustain. Energy Rev. 2025, 208, 115073. [ Google Scholar] [ CrossRef] Hudișteanu, V.-S.; Cherecheș, N.-C.; Țurcanu, F.-E.; Hudișteanu, I.; Romila, C. Impact of Temperature on the Efficiency of Monocrystalline and Polycrystalline Photovoltaic Panels: A Comprehensive Experimental Analysis for Sustainable Energy Solutions. Sustainability 2024, 16, 10566. [ Google Scholar] [ CrossRef] Abdul Rahim, M.S.B.; Tajuddin, M.F.N.B.; Saad, M.S.; Hasanuzzaman, M.; Ahmad, M.Z.Z.; Halim, A.A.A.; Mid, E.C. Experimental investigation on improving photovoltaic module efficiency using adaptive fuzzy-PID water cooling. Sol. Energy 2025, 300, 113828. [ Google Scholar] [ CrossRef] Sailor, D.J.; Anand, J.; King, R.R. Photovoltaics in the built environment: A critical review. Energy Build. 2021, 253, 111479. [ Google Scholar] [ CrossRef] He, Y.; Yu, H.; Zhao, M. Thermal Performance Study of Extensive Green Roof in Shanghai District: A Case Study of Lightweight Building in Winter. Procedia Eng. 2015, 121, 1597–1604. [ Google Scholar] [ CrossRef] Susca, T. Green roofs to reduce building energy use? A review on key structural factors of green roofs and their effects on urban climate. Build. Environ. 2019, 162, 106273. [ Google Scholar] [ CrossRef] Chabada, M.; Durica, P.; Juras, P. Experimental Analysis of the Influence of Seasonality on the Temperature Regime of Extensive Roofs in Central Europe. Buildings 2024, 14, 812. [ Google Scholar] [ CrossRef] Xu, H.; Chen, H.; Qian, C.; Li, J. The Evapotranspiration Characteristics and Evaporative Cooling Effects of Different Vegetation Types on an Intensive Green Roof: Dynamic Performance Under Different Weather Conditions. Sustainability 2024, 16, 10812. [ Google Scholar] [ CrossRef] Liu, C.; Zhang, X.; Jiang, M.; Zhu, S.; Wang, Z.; Long, J. The Coupled Thermal Response Analysis of Green Roofs Based on the Discrete Element Method. Buildings 2024, 14, 3984. [ Google Scholar] [ CrossRef] Juras, P. Rainwater Retention Test of Various Green Roofs: Influence on Membrane Temperatures and Evapotranspiration. Buildings 2023, 13, 2058. [ Google Scholar] [ CrossRef] Talwar, P.; Verma, N.; Khatri, H.; Ahire, P.D.; Chaudhary, G.; Lindenberger, C.; Vivekanand, V. A systematic review of photovoltaic-green roof systems in different climatic conditions focusing on sustainable cities and societies. Sustain. Cities Soc. 2023, 98, 104813. [ Google Scholar] [ CrossRef] Pitulac, O.; Chirilă, C.; Stătescu, F.; Marcoie, N. GIS-Based Assessment of Photovoltaic and Green Roof Potential in Iași, Romania. Appl. Sci. 2025, 15, 10786. [ Google Scholar] [ CrossRef] Zheng, Y.; Weng, Q. Modeling the Effect of Green Roof Systems and Photovoltaic Panels for Building Energy Savings to Mitigate Climate Change. Remote Sens. 2020, 12, 2402. [ Google Scholar] [ CrossRef] Yang, L.; Gao, X.; Li, Z.; Jia, D. Observational study on the thermal performance of photovoltaic and cool-photovoltaic roofs during heatwaves in a semi-arid city. Energy Build. 2026, 355, 117036. [ Google Scholar] [ CrossRef] Kazemian, A.; Xiang, C. Synergizing photovoltaic-thermal systems with green roofs: A pathway to enhanced urban sustainability and energy efficiency. Renew. Sustain. Energy Rev. 2025, 222, 116002. [ Google Scholar] [ CrossRef] Möslinger, L.; Streit, E.; Korjenic, A.; Sulejmanoski, A. Microclimatic Effects of Retrofitting a Green Roof Beneath an East–West PV Array: A Two-Year Field Study in Austria. Sustainability 2025, 17, 7495. [ Google Scholar] [ CrossRef] Fassbender, E.; Pytlik, S.; Rott, J.; Hemmerle, C. Impacts of Rooftop Photovoltaics on the Urban Thermal Microclimate: Metrological Investigations. Buildings 2023, 13, 2339. [ Google Scholar] [ CrossRef] Rahmatmand, A.; Harrison, S.J.; Oosthuizen, P.H. An innovative method to control surface temperature of a rooftop photovoltaic system. Sol. Energy 2020, 195, 581–591. [ Google Scholar] [ CrossRef] Soares de Moura, E.; Alves Marques, H.S.; Bressan Riffel, D. Assessment of photovoltaic I–V curve translation methods in real-world situations. Sol. Energy 2024, 276, 112654. [ Google Scholar] [ CrossRef] Appelbaum, J.; Maor, T. Dependence of PV Module Temperature on Incident Time-Dependent Solar Spectrum. Appl. Sci. 2020, 10, 914. [ Google Scholar] [ CrossRef] Lin, W.-T.; Chang, C.-M.; Huang, Y.-C.; Wu, C.-C.; Kuo, C.-C. Fault Diagnosis in Solar Array I-V Curves Using Characteristic Simulation and Multi-Input Models. Appl. Sci. 2024, 14, 5417. [ Google Scholar] [ CrossRef] Gulkowski, S.; Krawczak, E. Long-Term Energy Yield Analysis of the Rooftop PV System in Climate Conditions of Poland. Sustainability 2024, 16, 3348. [ Google Scholar] [ CrossRef] Jiang, H.; Zhang, X.; Yao, L.; Lu, N.; Qin, J.; Liu, T.; Zhou, C. High-resolution analysis of rooftop photovoltaic potential based on hourly generation simulations and load profiles. Appl. Energy 2023, 348, 121553. [ Google Scholar] [ CrossRef] Figure 1. Photograph of the experimental site showing the green roof and asphalt roof configurations with installed PV modules. Figure 1. Photograph of the experimental site showing the green roof and asphalt roof configurations with installed PV modules. Figure 2. Schematic side-view geometry of the experimental PV module installation showing module tilt angle and module-to-roof spacing. Figure 2. Schematic side-view geometry of the experimental PV module installation showing module tilt angle and module-to-roof spacing. Figure 3. The ALMEMO 500 high-precision datalogger (Ahlborn Mess-und Regelungstechnik GmbH, Holzkirchen, Germany) used for centralized monitoring and recording of environmental and thermal parameter. Figure 3. The ALMEMO 500 high-precision datalogger (Ahlborn Mess-und Regelungstechnik GmbH, Holzkirchen, Germany) used for centralized monitoring and recording of environmental and thermal parameter. Figure 4. The ZA9020-FS THERMO R2E4 high-precision temperature sensor (Ahlborn Mess-und Regelungstechnik GmbH, Holzkirchen, Germany) used for monitoring the temperature of the PV modules. Figure 4. The ZA9020-FS THERMO R2E4 high-precision temperature sensor (Ahlborn Mess-und Regelungstechnik GmbH, Holzkirchen, Germany) used for monitoring the temperature of the PV modules. Figure 5. Schematic illustration of the placement of temperature sensors on the rear side of the PV modules and on the roof surfaces. Sensors T1–T8 indicate the individual measurement points used for monitoring module and roof temperatures under asphalt and green roof conditions. Figure 5. Schematic illustration of the placement of temperature sensors on the rear side of the PV modules and on the roof surfaces. Sensors T1–T8 indicate the individual measurement points used for monitoring module and roof temperatures under asphalt and green roof conditions. Figure 6. The Testo 882 high-resolution thermal imager used for qualitative and quantitative thermal analysis of the PV module surfaces. Figure 6. The Testo 882 high-resolution thermal imager used for qualitative and quantitative thermal analysis of the PV module surfaces. Figure 7. Representative thermal images of PV modules installed above the asphalt roof ( a) and green roof ( b) surfaces captured under comparable operating conditions. Figure 7. Representative thermal images of PV modules installed above the asphalt roof ( a) an
