Global Horizontal Irradiance Research Articles

Ultra-short-term forecasting of global horizontal irradiance (GHI) integrating all-sky images and historical sequences

Accurate minute solar forecasts play an increasingly crucial role in achieving optimal intra-day power grid dispatch. However, continuous changes in cloud distribution and coverage pose a challenge to solar forecasting. This study presents a convolutional neural network-long short-term memory (CNN-LSTM) model to predict the future 10-min global horizontal irradiance (GHI) integrating all-sky image (ASI) and GHI sequences as input. The CNN is used to extract the sky features from ASI and a fully connected layer is used to extract historical GHI information. The resulting temporary information outputs are then merged and forwarded to the LSTM for forecasting the GHI values for the next 10 min. Compared to CNN solar radiation forecasting models, incorporating GHI into the forecasting process leads to an improvement of 18% in the accuracy of forecasting GHI values for the next 10 min. This improvement can be attributed to the inclusion of historical GHI sequences and regression via LSTM. The historical GHI contains valuable meteorological information such as aerosol optical thickness. In addition, the sensitivity analysis shows that the 1-lagged input length of the GHI and ASI sequence yields the most accurate forecasts. The advantages of CNN-LSTM facilitate power system stability and economic operation. Codes of the CNN-LSTM model in the public domain are available online on the GitHub repository https://github.com/zoey0919/CNN-LSTM-for-GHI-forecasting.

Explainable forecasting of global horizontal irradiance over multiple time steps using temporal fusion transformer

Accurate prediction of solar irradiance is essential for the successful integration of solar power plants into electrical systems. Despite recent advancements in deep learning technology yielding impressive results in solar forecasting, their lack of interpretability has hindered their widespread adoption. In this paper, we propose a novel approach that integrates a Temporal Fusion Transformer (TFT) with a McClear model to achieve accurate and interpretable forecasting performance. The TFT is a deep learning model that provides transparency in its predictions through the use of interpretable self-attention layers for long-term dependencies, recurrent layers for local processing, specialized components for feature selection, and gating layers to suppress extraneous components. The model is capable of learning temporal associations between continuous time-series variables, namely, historical global horizontal irradiance (GHI) and clear sky GHI, accounting for cloud cover variability and clear sky conditions that are often ignored by most machine learning solar forecasters. Additionally, it minimizes a quantile loss during training to produce accurate probabilistic forecasts. In this study, we evaluate the performance of hourly GHI forecasts on eight diverse datasets with varying climates: temperate, cold, arid, and equatorial, for multiple temporal horizons of 2, 3, 6, 12, and 24 h. The model is benchmarked against both climatological persistence for deterministic forecasting and Complete History Persistence Ensemble for probabilistic forecasting. To prove that our model is not location locked, it has been blind tested on four completely different datasets. The results demonstrate that the proposed model outperforms its counterparts across all forecast horizons.

Performance evaluation and comparative study of three 52-kW PV plants in India: a case study

Developing countries like India are rapidly transitioning from traditional energy sources to sustainable energy sources, due to the increase in demand and the depletion of fossil fuels. Grid-connected photovoltaic (PV) systems attract many investors, organizations, and institutions for deployment. This article studies and compares the performance evaluations of three 52-kW PV plants installed at an educational institution, SRMIST (SRM Institute of Science and Technology), in Tamil Nadu, India. This site receives an annual average temperature of 28.5°C and an average global horizontal irradiation of 160 kWh/m2/m. The prediction model for the 52-kW power plant is obtained using solar radiation, temperature, and wind speed. Linear regression model-based prediction equations are derived using the Minitab 16.2.1 software, and the results are compared with the real-time AC energy yield acquired from the three 52-kW plants for the year 2020. Furthermore, this 52-kW plant is designed using PVsyst V7.1.8 version software. The simulation results are compared with the energy yield from the plants in 2020 to identify the shortfall in the plant performance. The loss analysis for the plant is performed by obtaining the loss diagram from the PVsyst software. This study also proposes a methodology to study the commissioned PV plant's performance and determine the interaction between variables such as direct and diffused solar radiations, air temperature, and wind speed for forecasting hourly produced power. This article will motivate researchers to analyze installed power plants using modern technical tools.

Potential quantification and impact factors analysis of energy flexibility in residential buildings with preheating control strategies

Demand response (DR) is an effective method for achieving energy flexibility. Preheating can store a certain amount of energy in the building while simultaneously reducing the heat load. The energy flexibility of a building during the heating season mainly depends on the types of DR events, the thermal performance of the building envelope, and weather conditions. Analyzing the impact of these factors on the preheating performance and exploring their interactions can reveal the potential for energy flexibility. This study proposes a simulation-based method to evaluate the energy flexibility of a residential building in Kitakyushu, Japan from the perspectives of energy efficiency, load and energy reduction. Rule-based control strategies are applied to heating, ventilation, and air conditioning systems. The results show that compared to preheating temperature of 24 °C, high insulation building at 25 °C achieves energy efficiency improvement of 3.5%–14.8% and load reduction increase of 1.6%–22.2%. Compared to low insulation building, preheating at 25 °C for 1 h increases energy efficiency in high insulation building by 6.2%–10.3%, with corresponding load reduction increase of 1.0%–19.1%. Extending the preheating duration improves load reduction but leads to heat loss. Preheating during heating system startup should be avoided as it produces the worst energy flexibility. Moreover, statistical analysis identifies key indicators for energy reduction: the inside surface temperature of the exterior wall (Pearson correlation coefficient = 0.820, standard regression coefficient = 0.924), followed by global horizontal radiation and building insulation performance.

Diurnal Evolution and Estimates of Hourly Diffuse Radiation Based on Horizontal Global Radiation, in Cerrado-Amazon Transition, Brazil

In the Cerrado-Amazonian ecotone in the State of Mato Grosso, intensely altered by anthropic action, the knowledge and processes of energy conversion and energy balance are still incipient, making the monitoring and modeling of diffuse radiation essential for several applications. The objective of this study was to evaluate the seasonality of the diurnal evolution and estimate the hourly diffuse radiation (Hdh) and incident radiation in the horizontal plane between June 2011 and October 2016. The instantaneous measurements (5 min) of diffuse radiation underwent geometric, astronomical, and anisotropic corrections, with subsequent hourly integrations. The seasonality of diffuse radiation and its radiometric fractions was evaluated. The estimates were made considering total and seasonal data groupings (water stations in the region) and in different cloudiness classes (atmospheric transmissivity index—KTh). The diurnal behavior of diffuse radiation (Hdh) was similar to that of global radiation and at the top of the atmosphere, with maximum values at solar noon. The correlations between Kdh and KTh showed third-order polynomial behavior, with maximum observed values of Kdh ranging from 0.8 to 0.9, for KTh less than 0.2. Estimation equations based on radiometric fractions underestimated the values of diffuse radiation, with a better performance presented by models adjusted in annual data groupings. Among the parameterized models for estimating diffuse radiation obtained in the literature, those calibrated regionally in this study, together with those developed for tropical regions, presented better statistical performances.

A new accurate low-cost instrument for fast synchronized spatial measurements of light spectra

Abstract. We developed a cost-effective Fast-Response Optical Spectroscopy Time-synchronized instrument (FROST). FROST can measure 18 light spectra in 18 wavebands ranging from 400 to 950 nm with a 20 nm full-width half-maximum bandwidth. The FROST 10 Hz measurement frequency is time-synchronized by a global navigation satellite system (GNSS) timing pulse, and therefore multiple instruments can be deployed to measure spatial variation in solar radiation in perfect synchronization. We show that FROST is capable of measuring global horizontal irradiance (GHI) despite its limited spectral range. It is very capable of measuring photosynthetic active radiation (PAR) because 11 of its 18 wavebands are situated within the 400-to-700 nm range. A digital filter can be applied to these 11 wavebands to derive the photosynthetic photon flux density (PPFD) and retain information on the spectral composition of PAR. The 940 nm waveband can be used to derive information about atmospheric moisture. We showed that the silicon sensor has undetectable zero offsets for solar irradiance settings and that the temperature dependency as tested in an oven between 15 and 46 ∘C appears very low (−250 ppm K−1). For solar irradiance applications, the main uncertainty is caused by our polytetrafluoroethylene (PTFE) diffuser (Teflon), a common type of diffuser material for cosine-corrected spectral measurements. The oven experiments showed a significant jump in PTFE transmission of 2 % when increasing its temperature beyond 21 ∘C. The FROST total cost (< EUR 200) is much lower than that of current field spectroradiometers, PAR sensors, or pyranometers, and includes a mounting tripod, solar power supply, data logger and GNSS, and waterproof housing. FROST is a fully standalone measurement solution. It can be deployed anywhere with its own power supply and can be installed in vertical in-canopy profiles as well. This low cost makes it feasible to study spatial variation in solar irradiance using large-grid high-density sensor setups or to use FROST to replace existing PAR sensors for detailed spectral information.

CAELUS: Classification of sky conditions from 1-min time series of global solar irradiance using variability indices and dynamic thresholds

Precise sky classification as a function of cloudiness is desirable or necessary in a variety of applications. CAELUS, a novel classification algorithm that relies on various thresholds to separate all possible sky conditions into six classes, is presented here. It uses global horizontal irradiance (GHI) measurements at 1-min resolution, from which a set of four indices is derived to characterize the magnitude and temporal variability of GHI. The algorithm also requires precise estimates of 1-min GHI under hypothetical cloudless conditions, and the solar zenith angle (limited to a maximum of 85°). Using 1-min GHI measurements from 54 BSRN high-quality radiometric stations, which cover all five primary Köppen-Geiger climate classes, CAELUS is used here to classify their sky conditions. The classification results, including the distribution of sky classes and the transitions between consecutive sky classes, are found consistent with the known characteristics of each primary Köppen-Geiger climate. Moreover, in each climate class, the detection of 1-min cloudless situations is found comparable to that provided by two dedicated and state-of-the-art methods—Reno-Hansen and Bright-Sun.

Learning lessons from Murzuq-Libya meteorological station

In this study, an examination was conducted on weather data gathered from the Murzuq weather station over a period of nine months, specifically focusing on 15-minute time series solar radiation data. The data was sourced from the Center for Solar Energy Research and Studies in Tajoura-Tripoli, through a collaborative agreement between the Faculty of Engineering at Wadi Alshatti University and the research center. The information collected encompassed various solar radiation components, such as global horizontal solar radiation, direct normal radiation, sky-diffuse solar radiation, and ground reflected solar radiation. The aim of this study is to verify calculated values of these components using mathematical models by comparing them with their measured values. The investigation revealed that the Earth's reflectance value for the region was estimated and determined to be around 0.4. It is important to note that this figure was different from the typically advised value of 0.2 that was given in previous literature.

Prediction of Solar Irradiation in Africa using Linear-Nonlinear Hybrid Models

Solar irradiation prediction including Global Horizontal Irradiation (GHI) and Direct Normal Irradiation (DNI) is a useful technique for assessing the solar energy potential at specific locations. This study used five Artificial Neural Network (ANN) models and Multiple Linear Regression (MLR) to predict GHI and DNI in Africa. Additionally, a hybrid model combining MLR and ANNs was proposed to predict both GHI and DNI and improve the accuracy of individual ANN models. Solar radiation (GHI and DNI) and global meteorological data from 85 cities with different climatic conditions over Africa during 2001-2020 were used to train and test the models developed. The Pearson correlation coefficient was used to identify the most influential input variables to predict GHI and DNI. Two scenarios were proposed to achieve the goal, each with different input variables. The first scenario used influential input parameters, while the second incorporated geographical coordinates to assess their impact on solar radiation prediction accuracy. The results revealed that the suggested linear-nonlinear hybrid models outperformed all other models in terms of prediction accuracy. Moreover, the investigation revealed that geographical coordinates have a minimal impact on the prediction of solar radiation.

Smart Approaches for Evaluating Photosynthetically Active Radiation at Various Stations Based on MSG Prime Satellite Imagery

Photosynthetically active radiation (PAR) is the 400–700 nm portion of the solar radiation spectrum that photoautotrophic organisms including plants, algae, and cyanobacteria use for photosynthesis. PAR is a key variable in global ecosystem and Earth system modeling, playing a prominent role in carbon and water cycling. Alongside air temperature, water availability, and atmospheric CO2 concentration, PAR controls photosynthesis and consequently biomass productivity in general. The management of agricultural and horticultural crops, forests, grasslands, and even grasses at sports venues is a non-exhaustive list of applications for which an accurate knowledge of the PAR resource is desirable. Modern agrivoltaic systems also require a good knowledge of PAR in conjunction with the variables needed to monitor the co-located photovoltaic system. In situ quality-controlled PAR sensors provide high-quality information for specific locations. However, due to associated installation and maintenance costs, such high-quality data are relatively scarce and generally extend over a restricted and sometimes non-continuous period. Numerous studies have already demonstrated the potential offered by surface radiation estimates based on satellite information as reliable alternatives to in situ measurements. The accuracy of these estimations is site-dependent and is related, for example, to the local climate, landscape, and viewing angle of the satellite. To assess the accuracy of PAR satellite models, we inter-compared 11 methods for estimating 30 min surface PAR based on satellite-derived estimations at 33 ground-based station locations over several climate regions in Europe, Africa, and South America. Averaged across stations, the results showed average relative biases (relative to the measurement mean) across methods of 1 to 20%, an average relative standard deviation of 25 to 30%, an average relative root mean square error of 25% to 35% and a correlation coefficient always above 0.95 for all methods. Improved performance was seen for all methods at relatively cloud-free sites, and quality degraded towards the edge of the Meteosat Second Generation viewing area. A good compromise between computational time, memory allocation, and performance was achieved for most locations using the Jacovides coefficient applied to the global horizontal irradiance from HelioClim-3 or the CAMS Radiation Service. In conclusion, satellite estimations can provide a reliable alternative estimation of ground-based PAR for most applications.

Long-term implementation of a model predictive controller for a hydronic floor heating and cooling system in a highly glazed house in Canada

This article presents the results of a 182 day implementation of a model predictive controller (MPC) in a test house in Ottawa, Canada. The MPC controls the timing of hydronic floor heating and cooling within the highly glazed house (13.5% south-facing window-to-floor ratio). The objectives of the MPC are to optimally control the on/off times of active heating and cooling to minimize energy use while also minimizing air temperature violations outside of a prescribed range. It was found that 71% of the days in the experimental period saw successful operation of the MPC towards these goals. Days that experienced overheating were found to be concentrated in January–February and days that experienced overcooling were found to be concentrated in March–May. Problematic decision making by the MPC was found to be rooted in 4 main issues: (1) global horizontal irradiance (GHI) is a bad indicator of passive solar on its own; (2) weather forecasting (specifically GHI forecasting) is critical to MPC success in highly glazed buildings; (3) the GHI history regressor used in the prediction model regularly influenced predictions illogically; and (4) the MPC lacked decision-making logic on the timing of hydronic floor cooling.

Design and Development of Grid-Connected Solar PV Power Plant Using PVsyst

As we know, the biggest problem that threatens today's world is climate change and global warming. Alongside that, Afghanistan does not have a reliable source of power, and people who live in cities do not have full access to electricity. So, we need immediate access to electricity. Therefore, among all other renewable sources, the only one that is feasible in the target area and could solve this problem soon is investing in a solar PV power plant. This study aims to develop a standard procedure for the design of grid-connected solar PV systems using PVsyst software. The project began with a broad database of meteorological data, including global daily horizontal solar irradiance, and also a database of various renewable energy system components from different manufacturers. This paper will explain the grid solar power limited in the year 2023. The photovoltaic power plant has a solar radiation of 6.22 KWh/Sq./day, covering 162.66 acres of land. The operating module temperature varies from -40°C to 85°C, with a tilt angle of 32 degrees. The various power losses (PV losses due to irradiation level, temperature, soiling, inverter, wiring, power electronics, grid availability, and interconnection) amount to 2006 KWh/m2, with a total production of 60513314 KVAh/year, and the performance ratio is calculated from simulation, giving an annual PR of 75.16%.

On-Grid Mono-Si PV Solar System Performance Evaluation in Baghdad

Abstract The performance assessment of a 5 kWp On-Grid Mon-crystalline silicon photovoltaic( PV) solar system is the subject of the present paper. This PV system is located in Baghdad city, Al-Mansour company. The present PV solar system is observed during 2018. The current system’s yearly daily averages for the array yield final yield, and the reference yield are 4.65 kWh/kWp, 4.5 kWh/kWp, and 5.9 kWh/kWp, respectively. The final yield’s yearly energy is 1613.8 kWh/kWp. The present year’s electrical energy output from this system is 7891.57 kWh, and Baghdad’s annual global horizontal solar irradiation (GHI) is 1987.5 kWh/m2. The average yearly daily losses for the system and the array are 0.15 kWh/kWp and 1.33 kWh/kWp, respectively. The average yearly efficiencies for the system, the array and the inverter are 11.33%, 11.7%, and 96.75%, respectively. The annual daily averages for performance ratio and capacity factor are 74.7% and 17.75 %, respectively. Mono-crystalline silicon (Mono-Si) seems to operate well in the environment of Baghdad-Iraq.

Investigation of the effects of the Greek extreme wildfires of August 2021 on air quality and spectral solar irradiance

Abstract. In August 2021, a historic heatwave was recorded in Greece which resulted in extreme wildfire events that strongly affected the air quality over the city of Athens. Saharan dust was also transferred over Greece on certain days of the same period due to the prevailing southern winds. The impact of these events on air quality and surface solar radiation is investigated in this study. Event characterization based on active and passive remote sensing instrumentation has been performed. The study shows that significantly increased levels of air pollution were recorded from the end of July to the first week of August. The smoke led to unusually high aerosol optical depth (AOD) values (up to 3.6 at 500 nm), high Ångström exponent (AE) (up to 2.4 at 440–870 nm), and a strong and negative dependence of single-scattering albedo (SSA) on wavelength that was observed to decrease from 0.93 at 440 nm to 0.86 at 1020 nm, while the dust event led to high AOD (up to 0.7 at 500 nm), low AE (up to 0.9 at 440–870 nm), and a positive dependence of SSA on wavelength that was observed to increase from 0.89 at 440 nm to 0.95 at 1020. Furthermore, the smoke plume was also detected over the PANhellenic GEophysical observatory of Antikythera on 7 August, which is about 240 km away from Athens. Increased AOD values (up to ∼ 0.90 at 500 nm) associated with a high fine-mode AOD (up to ∼ 0.85 at 500 nm) and decrease in SSA with wavelength suggested the dominance of fine biomass burning aerosols. The impact of dust and smoke on solar irradiance revealed significant differences in the spectral dependence of the attenuation caused by the two different aerosol types. The attenuation of solar irradiance in the ultraviolet (UV-B) spectrum was found to be much lower in the case of dust compared to smoke for similar AOD500 values. Differences were less pronounced in the near-infrared and visible spectral regions. The large AODs during the wildfires resulted in a decrease in the noon UV index by up to 53 %, as well as in the daily effective doses for the production of vitamin D (up to 50 %), in the daily photosynthetically active radiation (up to 21 %) and in the daily global horizontal irradiance (up to 17 %), with serious implications for health, agriculture, and energy. This study highlights the wider impacts of wildfires that are part of the wider problem for Mediterranean countries, whose frequency is predicted to increase in view of the projected increasing occurrence of summer heatwaves.

Diffuse and Direct UV Index Experimental Values

In this work, we study the different components of the UV Index, direct, diffuse, and global, for seven years of UV erythemal measurements in a Mediterranean locality at Burjassot AtmoSpheric Station (BASS). The measurements have been taken with three YES-UVB-1 radiometers: one provided with a shadow disk on a Sun tracker; a radiometer to measure the global horizontal radiation; and a Sun-tracking radiometer to measure the direct radiation with an especially designed radiance collimator. It is known that diffuse UV irradiance represents at least 50% of total UV irradiance on a horizontal plane, which means being able to obtain a high UV Index even under shade. Our study aims to establish the relationship between the UV Index at noon and the maximum daily UV erythemal irradiance in the mid-latitude area, as well as the potential erythemal effects of exposure to solar UV radiation on different skin phototypes (skin phototypes I–IV) for both direct and diffuse components.

Hybrid Deep Learning Model for Mean Hourly Irradiance Probabilistic Forecasting

For grid stability, operation, and planning, solar irradiance forecasting is crucial. In this paper, we provide a method for predicting the Global Horizontal Irradiance (GHI) mean values one hour in advance. Sky images are utilized for training the various forecasting models along with measured meteorological data in order to account for the short-term variability of solar irradiance, which is mostly caused by the presence of clouds in the sky. Additionally, deep learning models like the multilayer perceptron (MLP), convolutional neural networks (CNN), long short-term memory (LSTM), or their hybridized forms are widely used for deterministic solar irradiance forecasting. The implementation of probabilistic solar irradiance forecasting, which is gaining prominence in grid management since it offers information on the likelihood of different outcomes, is another task we carry out using quantile regression. The novelty of this paper lies in the combination of a hybrid deep learning model (CNN-LSTM) with quantile regression for the computation of prediction intervals at different confidence levels. The training of the different machine learning algorithms is performed over a year’s worth of sky images and meteorological data from the years 2019 to 2020. The data were measured at the University of French Polynesia (17.5770° S, 149.6092° W), on the island of Tahiti, which has a tropical climate. Overall, the hybrid model (CNN-LSTM) is the best performing and most accurate in terms of deterministic and probabilistic metrics. In addition, it was found that the CNN, LSTM, and ANN show good results against persistence.

A novel approach suggestion for assessing the impact of topographic shading on the estimation of the floating photovoltaic technical potential

This study presents a novel approach for the implementation of floating photovoltaic (FPV) systems at the Ayvalı hydroelectric power plant (HPP) in Türkiye. The method proposed in this study accounts for dynamic changes in water levels to accurately calculate the shading effects induced by topography. First, the minimum reservoir surface for the FPV system was calculated using remote sensing (RS). The minimum reservoir surface area, which was determined as 504.69 ha using 60 Sentinel-2 satellite images, was calculated using machine learning algorithms on the Google Earth Engine (GEE) platform, support vector machines (SVM) and automatic water extraction index (AWEI). In the second stage, new digital elevation model (DEM) maps were produced by overlapping monthly changes in water height with ALOS PALSAR data and solar analysis was performed on them. An annual global horizontal irradiance (GHI) map was produced using these maps, and it was divided into five classes to emphasize differences in production potential. The results revealed that 1083.45 GWh of electricity can be produced annually by installing FPV in very high and high potential areas. However, as the moderate, low, and very low regions represent only 5.02% of the reservoir surface and there is a 1.68-fold difference in production potential between the highest and lowest areas due to topography-induced shading near the coastline, it was concluded that FPV installation would not be efficient in those regions. This study highlights the significance of incorporating topography-induced shading and emphasizes the importance of employing RS and geographic information system (GIS) techniques to achieve this objective.

Hourly global horizontal irradiance over West Africa: A case study of one-year satellite- and reanalysis-derived estimates vs. in situ measurements

Estimates of global horizontal irradiance (GHI) from reanalysis and satellite-based data are the most important information for the design and monitoring of PV systems in Africa, but their quality is unknown due to the lack of in situ measurements. In this study, we evaluate the performance of hourly GHI from state-of-the-art reanalysis and satellite-based products (ERA5, MERRA-2, CAMS, and SARAH-2) with 37 quality-controlled in situ measurements from novel meteorological networks established in Burkina Faso and Ghana under different weather conditions for the year 2020. The effects of clouds and aerosols are also considered in the analysis by using common performance measures for the main quality attributes and a new overall performance value for the joint assessment. The results show that satellite data performs better than reanalysis data under different atmospheric conditions. Nevertheless, both data sources exhibit significant bias of more than 150 W/m2 in terms of RMSE under cloudy skies compared to clear skies. The new measure of overall performance clearly shows that the hourly GHI derived from CAMS and SARAH-2 could serve as viable alternative data for assessing solar energy in the different climatic zones of West Africa.

Evaluation of the Solar Energy Nowcasting System (SENSE) during a 12-Months Intensive Measurement Campaign in Athens, Greece

Energy nowcasting is a valuable asset in managing energy loads and having real-time information on solar irradiation availability. In this study, we evaluate the spectrally integrated outputs of the SENSE system for solar irradiance nowcasting for the period of the ASPIRE (atmospheric parameters affecting spectral solar irradiance and solar energy) campaign (December 2020–December 2021) held in Athens, Greece. For the needs of the campaign, several ground-based instruments were operating, including two pyranometers, a pyrheliometer, a cloud camera, a CIMEL sunphotometer, and a precision spectral radiometer (PSR). Global horizontal irradiance (GHI) estimations were more accurate than direct normal irradiance (DNI). SENSE estimations are provided every 15 min, but when comparing bigger time intervals (hours-days), the statistics improved. A dedicated assessment of the SENSE’s inputs is performed in respect to ground-based retrievals, considering cloud conditions (from a sky imager), AOD, and precipitable water vapor from AERONET. The factor that established the larger errors was the visibility of the solar disc, which cannot be defined by the available sources of model inputs. Additionally, there were discrepancies between the satellite estimation of the clouds and the ground picture, which caused deviations in results. AOD differences affected more the DNI.

First estimation of high-resolution solar photovoltaic resource maps over China with Fengyun-4A satellite and machine learning

Fengyun-4A (FY-4A), which is the latest-generation Chinese geostationary meteorological satellite, measures solar reflection and thermal emission with high temporal, spatial, and spectral resolutions. It is expected to be highly beneficial for solar resource assessment and forecasting in China. This study is the first to estimate, using FY-4A and a random forest model, the global horizontal irradiance (GHI) at a 4-km–15-min spatio-temporal resolution over China, as a means to arrive at a solar photovoltaic (PV) resource map. In terms of GHI estimates, the root mean square error and mean bias error between hourly measured and retrieved values are 147.02 (35.2%), −5.64 W/m2 (−1.4%), respectively, whereas the values of daily estimates are 29.20 (18.0%), −2.97 W/m2 (−1.3%). The retrieval accuracy is found much better for instances with solar zenith angles smaller than 60°. Relatively larger errors are found at locations in the Sichuan Basin and northeastern China, which can be attributed to bright surfaces and/or strong cloud transients. With the retrieved irradiance, PV resource is derived through a physical model chain. The annual mean PV resource map suggests that, over most of the west areas, the annual mean effective irradiance exceeds 1700 kWh/m2, with the highest value found in Tibet (around 2000 kWh/m2 per annum). Eastern China has an annual effective irradiance of only 1300–1500 kWh/m2. The region with poorest solar resource is the Sichuan Basin (less than 1100 kWh/m2 per annum).