Extraterrestrial Solar Radiation Research Articles

Spatio-Temporal Deep Learning-Based Forecasting of Surface Solar Irradiance: Leveraging Satellite Data and Feature Selection

Spatio-Temporal Deep Learning-Based Forecasting of Surface Solar Irradiance: Leveraging Satellite Data and Feature Selection

Study of Global Horizontal and Vertical Solar Irradiance Components Under Different Sky Conditions for Sustainable Building Design In Bali Of Indonesia

The study of global and vertical solar irradiance components under different sky conditions is very important for building design applications. This research has been carried out by analyzing the components of solar radiation based on global horizontal irradiance (GHI) data, and global vertical irradiance for east, west, north, and south oriented. Observational data was acquired for one year starting from July 2022 to June 2023 in Jembrana-Bali of Indonesia. The solar radiation component has been analyzed based on sub-diurnal (daytime) and annual patterns. The clearness index (Kt) was calculated based on the ratio of extra-terrestrial solar radiation to global solar radiation measured at the surface. Sun-path analysis was carried out by considering the relationship of the zenith angle, azimuth angle, and the intensity of each solar radiation component to changes in time. The results show that in all-sky conditions, the average monthly maximum global horizontal irradiance occurs in January at 432.4 W/m², while the minimum monthly average global horizontal irradiance occurs in July at 327.9 W/m². The clearness index statistic shows that the hourly average ranges from 0.12 to 0.68 with an average of 0.47. Based on sun-path analysis, east, north, and west orientations receive more solar radiation in the Jembrana-Bali area. The findings from the study provide important information for architects, engineers, and policymakers that can be used for sustainable building design and building energy planning.

Performance Analysis of Solar Water Heater – A Case Study in Al Seeb, Oman

The 2040 vision of the Sultanate of Oman has given remarkable attention to providing green energy supplies to different sectors in the country. The solar water heater system is projected to boost Oman's renewable energy utilisation and help achieve the country’s energy sustainability. Due to extreme weather conditions, electric water heaters consume most of the electricity, mainly during the winter. In this study, the evacuated tube solar water heater is installed, and the system's performance at different inclination angles and weather conditions is investigated. The experiment was conducted during various weather conditions and at three different inclination angles, such as 45°, 56° and 65°. It was observed that a temperature close to 100°C was achieved during the daytime. The performance analysis of the evacuated solar water heater shows the water temperature was still high even during the evening. Extraterrestrial solar radiation plays a vital role in maintaining the water temperature in the evening. Moreover, an inclination angle of 45° performed well and produced a higher temperature of more than 90°C.Keywords: Solar Thermal, solar collector, solar water heater, temperature, performance.

Evapotranspiration Estimation Using Machine Learning Methods

Abstract The study examined the performance of four machine learning algorithms (regression trees, boosted trees, random forests, and artificial neural networks) for estimating evapotranspiration (ETo) based on incomplete meteorological data. Meteorological variables (mean and maximum air temperature, average air humidity, average level of solar radiation, vapor pressure deficit, extraterrestrial solar radiation, and day number of the year) were used as input. The simulation used two calculation scenarios: data with and without average solar radiation. The performance of the different machine learning models was evaluated using the mean square error, root mean square error, coefficient of determination, and slope of regression forced through the origin between the measured and simulated ETo. The results demonstrated that the applied models were able to describe nonlinear relationships between weather parameters and evapotranspiration. The accuracy of evapotranspiration estimation depended on the type of input variables and the machine learning model used. The highest level of evapotranspiration prediction was obtained using the artificial neural networks model. Including solar radiation data in the calculations improved the quality of evapotranspiration prediction in all four models. In the absence of data on the actual solar radiation reaching the Earth's surface, it is advisable to supplement the input data with data on extraterrestrial solar radiation and the day number of the year. Such an approach can be helpful in areas and situations with limited access to meteorological data.

Reference evapotranspiration projections in Southern Spain (until 2100) using temperature-based machine learning models

This study firstly examines the performance of temperature-based machine learning models for estimating reference evapotranspiration (ET0), an essential parameter for water management in agriculture, ecosystems, and hydrology. Data from 122 Automated Weather Stations (AWS) across different regions in Southern Spain has been studied and four machine learning models have been developed and assessed: Multilayer Perceptron (MLP), Extreme Learning Machine (ELM), Random Forest (RF), and Support Vector Machine (SVM). The results indicate that all machine learning models outperform the traditional Hargreaves-Samani method in estimating ET0. Specifically, ELM performed, on average, as the best in terms of Global Performance Indicator (GPI) for those locations situated in the first region (GPI = 0.1860), while MLP outperformed the rest for those located in the second (GPI = 0,1162). Besides, the configuration using minimum, mean and maximum air temperature (Tx, Tm, Tn, respectively), the diurnal temperature range (DTR), and Extraterrestrial solar radiation features (Ra) was found to be the fittest for the second region (GPI = 0.0734) and that using Tx, Tn, Tm and Ra in the first one in (GPI = 0.1938). Once the models were validated, they were applied to future 5 km gridded projection datasets, using different Representative Concentration Pathway (RCP) scenarios, in order to estimate ET0 up to the year 2100. In general, the projected ET0 was found to increase significantly in the future, with Mann Kendall Z values that ranged from 7.11 to 10.37 in the RCP4.5 scenario and from 10.84 to 12.57 in the RCP8.5 scenario. Thus, the ET0 is expected to increase from 1300 to 1600 mm to 1500–1700 mm using the RCP4.5 and to 1900 mm using the RCP8.5 in Andalusia, with the highest increase occurring in the south coastal region. This study provides important insights into the application of machine learning models to estimate ET0 and its implications for future water management strategies.

Study and characterization of global solar radiation and incident UV in the city of Leon, Guanajuato, Mexico

Controlled sun exposure can greatly benefit vitamin D synthesis, however, irresponsible sun exposure can cause considerable biological damage. Among the extraterrestrial solar radiation that reaches the earth's surface, UV rays are characterized by having the shortest wavelengths (less than 400 nm) and the highest frequencies, which is why they concentrate a large amount of energy. In particular, radiation with wavelengths less than 300 nm can alter DNA molecules and cause biological damage such as skin changes, cataracts, possible mutations, immunosuppression, photoaging and many other conditions. Therefore, this research estimates the global solar radiation and the incident daily UV radiation, as well as the UV index in order to identify the possible dose of absorbed radiation and consequently identify the relevant protection actions for the local population.

Simplified Method for Predicting Hourly Global Solar Radiation Using Extraterrestrial Radiation and Limited Weather Forecast Parameters

Solar radiation has important impacts on buildings such as for cooling/heating load forecasting, energy consumption forecasting, and multi-energy complementary optimization. Two types of solar radiation data are commonly used in buildings: radiation data in typical meteorological years and measured radiation data from meteorological stations, both of which are types of historical data. However, it is difficult to predict the hourly global solar radiation, which affects the application of relevant prediction models in practical engineering. Most existing methods for predicting hourly global solar radiation have issues such as difficulty in obtaining input parameters or complex data processing, which limits their practical engineering applications. This study proposed a simplified method to accurately predict the hourly horizontal solar radiation using extraterrestrial solar radiation, weather types, cloud cover, air temperature, relative humidity, and time as the input parameters. The back-propagation network, support vector machine, and light gradient boosting machine (LightGBM) models were used to establish the prediction model, and Shapley additive explanations were used to analyze the relationship between the input variables and the prediction results to simplify the structure of the prediction model. Taking Lanzhou New District in Gansu Province as an example, the results showed that the LightGBM model performed the best, with the root mean square error of 126.1 W/m2. Shapley additive explanations analysis showed that weather type was not a significant factor in the LightGBM model. Therefore, the weather type was removed from the LightGBM model and the root mean square error was 135.2 W/m2. The results showed that extra-terrestrial radiation and limited weather forecast parameters can be used to predict hourly global solar radiation with satisfactory prediction results.

Application of Exergy for Research on Increasing the Usefulness of Solar Radiation by Dispersing It into Monochromatic Beams

Energy determines the ability of matter to work. However, in the given environment, the real usefulness to perform work is determined by exergy. This study covers not only solar, but also any monochromatic thermal radiation. The value of such radiation was determined by its exergy and the ratio of its exergy-to-energy. A novelty in this work is to demonstrate by means of exergy that the usefulness of thermal polychromatic radiation can be increased by its dispersion to monochromatic radiation. This effect is the greater, the lower the temperature of the radiation. Analogies of this effect to the exergetic effect of gas separation have been indicated. The effect of the increase in exergy in the process of radiation dispersion was interpreted by means of a cylinder-piston system that explains this effect with the influence of environmental radiation. The concept of quasi-monochromatic and cumulated radiation was introduced into dispersion considerations and the change in the energetic, entropic and environmental components of the exergy of radiation beams was analyzed. Considerations were illustrated with appropriate examples of calculations considering dispersion of high-temperature radiation, such as extraterrestrial solar radiation and dispersion of low-temperature radiation from water vapor.

Data-Driven Models for Predicting Solar Radiation in Semi-Arid Regions

Solar energy represents one of the most important renewable energy sources contributing to the energy transition process. Considering that the observation of daily global solar radiation (GSR) is not affordable in some parts of the globe, there is an imperative need to develop alternative ways to predict it. Therefore, the main objective of this study is to evaluate the performance of different hybrid data-driven techniques in predicting daily GSR in semi-arid regions, such as the majority of Spanish territory. Here, four ensemble-based hybrid models were developed by hybridizing Additive Regression (AR) with Random Forest (RF), Locally Weighted Linear Regression (LWLR), Random Subspace (RS), and M5P. The base algorithms of the developed models are scarcely applied in previous studies to predict solar radiation. The testing phase outcomes demonstrated that the AR-RF models outperform all other hybrid models. The provided models were validated by statistical metrics, such as the correlation coefficient (R) and root mean square error (RMSE). The results proved that Scenario #6, utilizing extraterrestrial solar radiation, relative humidity, wind speed, and mean, maximum, and minimum ambient air temperatures as the model inputs, leads to the most accurate predictions among all scenarios (R = 0.968–0.988 and RMSE = 1.274–1.403 MJ/m2⋅d). Also, Scenario #3 stood in the next rank of accuracy for predicting the solar radiation in both validating stations. The AD-RF model was the best predictive, followed by AD-RS and AD-LWLR. Hence, this study recommends new effective methods to predict GSR in semi-arid regions.

Effect of the net radiation proxies on maize and soya evapotranspiration estimation using machine learning methods.

Accurate evapotranspiration (ET) estimation is essential for water management in crops, but it is not an easy task. Empirical ET methodologies require precise net radiation (Rn) measurements to obtain accurate results. Nevertheless, Rn measurements are not easy to obtain from meteorological stations. Thus, this study explored the use of machine learning algorithms with two Rn substitutes, to estimate daily ET: the extraterrestrial solar radiation (Ra) and a modelled Rn (RnM). Support Vector Machine (SVM), Kernel Ridge (KR), Decision Tree (DT), Adaptive Boosting (AB), and Multilayer Perceptron (MLP) were applied to model FLUXNET Rn and ET observations. Adaptive Boosting produced the best field Rn measurements (RnO), yielding a Root Mean Square Error of about 16 % of the mean observed Rn. The resulting Rn (AB RnM) was used to model daily crops ET employing the above-mentioned machine learning methods with RnO, AB RnM, and Ra, in conjunction with meteorological variables and the NDVI index. The evaluated methods were suitable to estimate ET, yielding similar errors to those obtained with RnO, when contrasted with ET observations. These results demonstrate that AB and KR are applicable with rutinary meteorological and satellite data to estimate ET.

Short-term solar radiation forecasting with a novel image processing-based deep learning approach

— In this study, an image processing-based deep learning approach for short-term forecast of solar radiation has been developed. For this purpose, firstly, cloud movements occurred during the day are tracked and future cloud movements are forecasted, accordingly. Subsequently, using the cloud motion estimation and extraterrestrial solar radiation data, 1-min averaged solar radiation values are estimated for 5-min time horizon. Shi-Tomasi method is employed to determine the feature points to be tracked on the sky images whereas, Lucas-Kanade optical flow method is employed to track the determined feature points on the sequential images. Average cloud velocity and directions are calculated by the help of linear regression method from tracked cloud movements. A hybrid approach including K-means and red/blue ratio is built to classify the pixels of the image whether they are clouds or sky. Finally, short-term solar radiations are estimated using the Long-Short Term Memory (LSTM) deep learning method. The performance of the proposed approach is compared with other methods in the literature. As a result it is concluded that, developed approach outperforms most methods in the literature with RMSE values of 47.576, 53.830, 68.103, and 92.386 for four different days and can be used as an alternative approach.

Developing machine learning models for air temperature estimation using MODIS data

Air temperature is a key variable in a wide range of environmental applications, including land–atmosphere interaction, climate change research and hydrology and crop growth models, among others. The objective of this study was to estimate daily maximum (Tmax) and minimum (Tmin) temperatures, based on MODIS AQUA/TERRA land surface temperature (LST), NDVI, extraterrestrial solar radiation and precipitation data. Artificial neural networks (ANN) and random forests (RF) models were developed to predict these temperatures covering weather stations in Córdoba (Argentina) for 2018-2020. The results show that RF and ANN machine learning algorithms are capable of modeling non-linear relationships between registered temperatures and LST MODIS data, in a very robust way. The validation of the models confirms that Tmax and Tmin can be accurately estimated using, jointly or separately, AQUA and TERRA LST. The best models present determination coefficients equal to 0.81/0.91 and root mean square error of 2.7/2.1 ºC for Tmax/Tmin, when using AQUA LST day/night satellite overpass time data, respectively. The robustness and confidence of the models developed, and the ease and free accessibility of input data at a global scale, suggest that these methodologies have the potential to be applied to other regions.

Demonstration of Net Solar Radiation Geographical Behavior Revers Correlation with Relative Humidity in Iraq

It is common that extraterrestrial solar radiation Ra value lessen due to the interaction with atmosphere components to result a net solar radiation Rn value about one third of Ra for most measurements on earth. Net solar radiation had major impact on air temperature (To) value statement. So, its estimation is crucial for analyzing the interaction of (atmosphere-vegetation-soil). Many methodologies available for this aim directly (with radiometer) or indirectly (with air temperature-based formula or soil plus air temperatures-based formula). The peak influences of global warming phenomenon were in eighties decade of the last century, yielding Iraq Rn rates to get climax after that Rn rates behaved in geographical based style, so those rates raised in middle and south regions of Iraq, and descending in its northern parts. The reason (as demonstrated in this research) was the spatial correlation in a reverse manner with RH rates all over Iraq.

Classification of Sky Cover by the Clearness Index (Kt) in Maputo - Mozambique

The present paper makes an analysis of the frequency of sky conditions in the city of Maputo/Mozambique based on the clearness index (Kt). From the determination of extraterrestrial solar radiation and measured data of four years of global daily radiation, Kt was calculated. Different sky conditions were classified according to the Escobedo criterion, which dispenses the data of direct and / or diffuse radiation in its classification. The city of Maputo, during the year, is observed more days with conditions of clear sky, in average of 120 days. And for cloudy sky conditions, the city of Maputo presents the least number of days, with an average of 48 days. The frequency of days for partially cloudy sky and partially clear sky conditions is 85 and 112 days respectively. The information generated in this paper can be used in future studies to project solar systems for thermal use or direct conversion of solar radiation into electricity.

A Scheme for Quickly Simulating Extraterrestrial Solar Radiation over Complex Terrain on a Large Spatial-Temporal Span—A Case Study over the Entirety of China

Extraterrestrial solar radiation (ESR) is the essential basic background for solar radiation, which determines the occurrence of the weather and atmospheric phenomena. Since the influence of ESR variation on actual rugged terrain is a diverse, complex, and dynamic process, simulating ESR over a large spatial-temporal span, especially with a high-resolution digital elevation model (DEM), is a significant challenge. In this paper, we developed a new scheme for simulating ESR over the entirety of China using a DEM with a resolution of 30 m. To fully consider regional terrain status, the feature variables used were elevation, slope, and aspects of the located grid and the surrounding four grids to reveal the topography. In addition, latitude was used as a feature variable to consider the geographical location, and the month number was used to consider the duration. On the basis of different geographical locations, the training dataset was established from 20,000 grids. With the feature variable composition and training dataset, a backpropagation artificial neural network (BP ANN) was found to have the best performance compared with the other three machine learning methods in simulating ESR for a DEM. In terms of the proposed scheme and BP ANN, we drew an ESR map of China with a resolution of 30 m. The determination coefficient of the simulation result achieved 0.99 and the root-mean-square error was less than 50 MJ/m2 in all sample areas, confirming its remarkable accuracy. In terms of efficiency, the time consumption of ESR simulated using the proposed scheme shrinks over 150 times in all sample areas compared to that simulated via the theoretical model. Simultaneously, the developed scheme was also used to simulate an ESR for a DEM with a resolution of 90 m to verify the universality and robustness of the developed scheme. In addition, we used the proposed scheme to derive the direct solar radiation and global solar radiation, thereby further proving the reliability and applicability of our study. Overall, our work convincingly proved that the proposed scheme is a potential and effective approach for quickly simulating ESR with high accuracy. This study provides the basis for different solar radiation inversions of long time series and large spatial scales, offering additional insights for simulating ESR on a large spatial-temporal span.

Machine Learning Models Based on Random Forest Feature Selection and Bayesian Optimization for Predicting Daily Global Solar Radiation

Prediction of daily global solar radiation with simple and highly accurate models would be beneficial for solar energy conversion systems. In this paper, we proposed a hybrid machine learning methodology integrating two feature selection methods and a Bayesian optimization algorithm to predict H in the city of Fez, Morocco. First, we identified the most significant predictors using two Random Forest methods of feature importance: Mean Decrease in Impurity (MDI) and Mean Decrease in Accuracy (MDA). Then, based on the feature selection results, ten models were developed and compared: (1) five standalone machine learning (ML) models including Classification and Regression Trees (CART), Random Forests (RF), Bagged Trees Regression (BTR), Support Vector Regression (SVR), and Multi-Layer Perceptron (MLP); and (2) the same models tuned by the Bayesian optimization (BO) algorithm: CART-BO, RF-BO, BTR-BO, SVR-BO, and MLP-BO. Both MDI and MDA techniques revealed that extraterrestrial solar radiation and sunshine duration fraction were the most influential features. The BO approach improved the predictive accuracy of MLP, CART, SVR, and BTR models and prevented the CART model from overfitting. The best improvements were obtained using the MLP model, where RMSE and MAE were reduced by 17.6% and 17.2%, respectively. Among the studied models, the SVR-BO algorithm provided the best trade-off between prediction accuracy (RMSE=0.4473kWh/m²/day, MAE=0.3381kWh/m²/day, and R²=0.9465), stability (with a 0.0033kWh/m²/day increase in RMSE), and computational cost.

Effects of clouds and aerosols on ecosystem exchange, water and light use efficiency in a humid region orchard

Investigating the patterns of water and carbon dynamics in agro-ecosystems in response to clouds and aerosols can shed new insights in understanding the biophysical impacts of climate change on crop productivity and water consumption. In this study, the effects of clouds and aerosols as well as other environmental factors on ecosystem water and carbon fluxes were examined based on three-year eddy covariance measurements under different sky conditions (quantified as the clearness index, Kt, i.e., the ratio of global solar radiation to extraterrestrial solar radiation) in a kiwifruit plantation in the humid Sichuan Basin of China. Results showed that evapotranspiration (ET) and canopy transpiration (Tc, measured by sap flow sensors) increased, while ecosystem light use efficiency (eLUE) and ecosystem water use efficiency (eWUE) decreased with increasing Kt. GPP presented a parabolic relationship with increasing Kt. The path analysis revealed that surface conductance (Gs) and canopy conductance (Gc) were the most dominant variables directly regulated carbon (GPP) and water (ET and Tc) fluxes. The effect path of Kt on ET and Tc was converted from through diffuse photosynthetic active radiation (PARdif) to direct PAR (PARdir) when the sky became clearer. The effect path of Kt on GPP was primarily through PARdif under different sky conditions. The declined eWUE with increasing Kt was caused by the different responses of GPP and ET to PARdir under clear skies. The declined eLUE resulted from the sharp decrease in GPP/PARdir, which surpassed the slight increase of GPP/PARdif with increasing PAR. The Priestley-Taylor Jet Propulsion Laboratory ET model (PT-JPL) incorporating Kt with an exponential function produced more reliable Tc estimates but minor improvement in ET. Further, the LUE-GPP model incorporating Kt with a linear function obtained much better GPP estimates. Our study shed light on how sky conditions modulate water and carbon dynamics between the biosphere and atmosphere, highlighting the necessity of the inclusion of sky conditions for better modeling regional water and carbon budgets.

Estimation of evapotranspiration in Eucalyptus plantation and mixed forests based on air temperature and humidity

Soil evaporation, Air temperature and humidity inside and outside the forest and river flows were continually monitored in eucalyptus plantations (EU) and coniferous broadleaved mixed forests (MFs) to investigate the reasons for the decreased runoff recorded in EU. The applicabilities of four estimation methods based on air temperature and humidity were evaluated and verified. Our results show that: (1) the average annual solar radiation penetration rate of the EU canopies was 1.75 times that of the MFs canopies, and temperatures were lower inside the two forests than outside the forests. The daily average temperatures inside the EU and MFs were 20.7 ℃ and 20.8 ℃, respectively, which were 1.5 ℃ and 1.4 ℃ lower than those outside the forests. The daily average relative humidity was 0.8% higher in the MFs than in the EU. (2) The evapotranspiration of EU and MFs accounted for 80.2% and 76.3% of the total rainfall respectively, of which the soil evaporation was 12.8% and 8.9% respectively, and the runoff production coefficients were 0.198 and 0.237 respectively. The main reason for the decrease of the EU production was that soil evaporation in EU was 43.8% higher than that in MFs. (3) The deviation of EU and MFs evapotranspiration (ET0) from the measured values was estimated by the ET0Valian(T, RH) and ET0Alt/rs(T, RH) methods as less than 10%. (4) Using the product of the extraterrestrial solar radiation and the surface solar emissivity of the forest instead of the extraterrestrial solar radiation in these four formulas, the variation trend of the estimated forest soil evaporation was highly similar to the measured value. ET0Valian(T, RH) and ET0Alt/rs(T, RH) had the highest accuracy in soil evaporation estimation (2.0–3.4% lower than the measured values). The results provide a theoretical basis for studying the ecohydrology of subtropical plantations in the Northern Hemisphere.

Estimating reference evapotranspiration using hybrid adaptive fuzzy inferencing coupled with heuristic algorithms

Hybrid heuristic algorithm (HA), an innovative technique in the machine learning field, enhances the accuracy of reference evapotranspiration (ETo) prediction, which is of paramount significance for regional water management, agricultural planning, and irrigation designing. However, the new hybrid HA techniques, namely Moth-Flame Optimization Algorithm (MFO) and Water Cycle Optimization Algorithm (WCA) are rarely applied to estimate ETo in the earlier literature. Therefore, this study assessed prediction and the estimation abilities of a novel hybrid adaptive neuro-fuzzy inference system (ANFIS-WCAMFO) for monthly ETo of Dhaka and Mymensing stations with data-limited humid regions of south-central Bangladesh. Prediction precision of the ANFIS-WCAMFO model is compared with other state-of-art models, i.e., ANFIS-WCA and ANFIS-MFO using a 4-fold cross-validation method including root-mean-square-error (RMSE), mean absolute error (MAE), and Nash-Sutcliffe efficiency (NSE), and coefficient of determination (R2). Nine input combinations of meteorological datasets, including extraterrestrial radiation (Ra), solar radiation (Rs), maximum and minimum temperatures (Tmax and Tmin), relative humidity (RH), and wind speed (U2), were employed for model training and testing purposes. The ANFIS-WCAMFO performed superior to the other state-of-arts methods in estimating monthly ETo in all input combinations. The ANFIS-WCA, ANFIS-MFO, and ANFIS-WCAMFO hybrid models for estimating ETo improved RMSE as 2.7%, 6.9%, and 15.1% for Dhaka station and 0.6%, 7.3%, 12.4% for Mymensingh station, respectively. The use of the Ra variable with the temperature inputs considerably improved the models’ accuracy in ETo; improvements in RMSE, MAE, NSE, and R2 of the hybrid ANFIS-WCMFO models were 26.6%, 30.9%, 17.6%, and 8.2% for Dhaka station and by 28.8%, 34.2%, 18.8% and 22.1% for Mymensingh station, respectively. The proposed hybrid neuro-fuzzy model has been suggested as a promising technique due to high predictive accuracy and less error for monthly ETo prediction in a data-limited tropical humid region.

An Interpretable Machine Learning Model for Daily Global Solar Radiation Prediction

Machine learning (ML) models are commonly used in solar modeling due to their high predictive accuracy. However, the predictions of these models are difficult to explain and trust. This paper aims to demonstrate the utility of two interpretation techniques to explain and improve the predictions of ML models. We compared first the predictive performance of Light Gradient Boosting (LightGBM) with three benchmark models, including multilayer perceptron (MLP), multiple linear regression (MLR), and support-vector regression (SVR), for estimating the global solar radiation (H) in the city of Fez, Morocco. Then, the predictions of the most accurate model were explained by two model-agnostic explanation techniques: permutation feature importance (PFI) and Shapley additive explanations (SHAP). The results indicated that LightGBM (R2 = 0.9377, RMSE = 0.4827 kWh/m2, MAE = 0.3614 kWh/m2) provides similar predictive accuracy as SVR, and outperformed MLP and MLR in the testing stage. Both PFI and SHAP methods showed that extraterrestrial solar radiation (H0) and sunshine duration fraction (SF) are the two most important parameters that affect H estimation. Moreover, the SHAP method established how each feature influences the LightGBM estimations. The predictive accuracy of the LightGBM model was further improved slightly after re-examination of features, where the model combining H0, SF, and RH was better than the model with all features.