Annals of the Academy of Romanian Scientists

Series on Engineering Sciences

16

Volume 18 Number 1/2026

ISSN 2066-6950

GEOSPATIAL MODELLING OF SOLAR RADIATION

USING SATELLITE-DERIVED DATA

Marina ANTONESCU¹,

Georgeta BANDOC²

Rezumat. Studiul analizează distribuția spațială și variabilitatea interanuală a resursei

solare în România prin utilizarea seriilor temporale orare derivate din date satelitare și a

modelării geostatistice în mediu GIS. Analiza vizează principalele componente ale

bilanțului radiativ la suprafață — radiația globală pe plan orizontal (GHI), radiația

directă normală (DNI), radiația difuză pe plan orizontal (DHI) și componenta directă pe

plan orizontal (BHI), derivată din DNI — calculate pentru intervale sezoniere și anuale

pentru perioada 2004–2025. În contextul disponibilității reduse a observațiilor

radiometrice la sol, integrarea datelor satelitare multianuale cu instrumente GIS și tehnici

de modelare geostatistică permite o analiză integrată a configurației spațiale și a

variabilității interanuale a radiației solare în România și oferă, totodată, un cadru analitic

pentru evaluări ulterioare ale potențialului energetic solar.

Abstract. This study analyzes the spatial distribution and interannual variability of the

solar resource in Romania using hourly time series derived from satellite data and

geostatistical modeling in a GIS environment. The analysis focuses on the main components

of the surface radiation balance—global horizontal irradiance (GHI), direct normal

irradiance (DNI), diffuse horizontal irradiance (DHI), and the direct horizontal component

(BHI), derived from DNI—calculated at seasonal and annual scales for the period 2004–

2025. Given the limited availability of ground-based radiometric observations, the

integration of long-term satellite data with GIS tools and geostatistical modeling

techniques enables an integrated analysis of the spatial patterns and interannual variability

of solar radiation in Romania, while also providing an analytical framework for subsequent

assessments of solar energy potential.

Keywords: solar resource assessment, satellite-derived irradiance, geostatistical

interpolation, interannual variability, Romania.

DOI

10.56082/annalsarscieng.2026.1.16

1. Introduction

The global transition towards low-carbon energy systems has intensified

research into the identification, quantification, and effective utilisation of renewable

energy resources. Within this broader context, solar energy has assumed a central

¹PhD (ABD), Faculty of Geography, University of Bucharest, Bucharest, Romania; Meteo Romania

(National Meteorological Administration), Bucharest, Romania; (marina_antonescu@yahoo.com).

²Prof., PhD, Faculty of Geography, University of Bucharest, Bucharest, Romania; Academy of

Romanian Scientists, Bucharest, Romania; (bandoc@geo.unibuc.ro).

Geospatial modelling of solar radiation using satellite-derived data

17

role, owing to its abundance and widespread availability at the Earth’s surface [1].

The energy potential associated with incoming solar radiation greatly exceeds

current global energy demand [1]. Its effective exploitation, however, depends on

a rigorous characterisation of the spatial distribution and temporal variability of the

main radiative parameters [2].

Solar resource assessment represents a fundamental phase in the planning

and design of both photovoltaic and concentrating solar power systems, as the

technical performance and economic viability of such installations are directly

conditioned by the level and temporal stability of incident radiation [3].

Solar resource assessment has long relied on ground-based radiometric

observations, which are still regarded as the reference source for solar resource

characterisation [4,5]. Nevertheless, the sparse density and uneven spatial

distribution of measurement stations often limit the coherent description of solar

resources at regional and national scales [6,7,8]. These constraints led to the

widespread adoption of solar irradiance products derived from satellite observations

and atmospheric modelling, which provide continuous spatial coverage and

temporally consistent records over extended periods [6,9]. Such products, however,

remain subject to uncertainty, as their performance is influenced by the

representation of clouds and aerosols, as well as by the radiative parameterisations

embedded in the underlying models, potentially leading to systematic regional or

seasonal biases [10,11]. Despite these limitations, satellite-derived time series

enable robust multiannual assessments with high spatial comparability, thereby

supporting the identification of the climatological characteristics of solar radiation

and the analysis of its interannual variability [12,13].

Beyond data availability, another important aspect of solar resource

assessment lies in the integration and spatial analysis of radiative information in

relation to geographical and topographical factors. Geographic Information

Systems (GIS) provide the technological framework required for the processing,

modelling, and spatial representation of continuous climatic variables, enabling the

integration of satellite-derived data with topographic, administrative, and

infrastructural information [14,15]. Numerous studies have demonstrated the utility

of GIS for mapping radiative parameters, analysing the influence of topography on

their spatial distribution, and assessing solar energy potential at regional scales

[14,16,17]. Through the application of geostatistical techniques, GIS also enables

the estimation of the spatial distribution of radiation in areas lacking direct

observations, while accounting for spatial autocorrelation and the influence of

explanatory variables.

Within this context, the integration of hourly satellite-derived solar radiation

data within a GIS framework [18], combined with the statistical analysis of

interannual variability [19], provides a robust analytical basis for solar resource

assessment at the national scale. Such an approach enables not only the

18

Marina Antonescu, Georgeta Bandoc

quantification of mean radiation levels, but also the evaluation of their temporal

stability, an aspect that is essential for energy planning and for assessing the risks

associated with climatic variability.

Against this background, the present study examines the spatial distribution

and seasonal variability of solar radiation in Romania using satellite-derived hourly

data for the 2004–2025 period together with geostatistical modelling within a GIS

framework. The analysis relates the geospatial organisation of the principal

radiative components and the statistical expression of their variability, with the aim

of evaluating the solar resource at national scale.

2. Data and Methodology

2.1.

Satellite-Derived Solar Radiation Dataset

The solar radiation data used in this study were obtained from the CAMS

Solar Radiation Time Series service, developed within the Copernicus programme

under the Copernicus Atmosphere Monitoring Service (CAMS) and accessed

through the Atmosphere Data Store (ADS) [20]. This product provides hourly

estimates of surface radiative parameters derived from the integration of satellite

observations with numerical atmospheric models, thereby ensuring temporal

consistency and continuous spatial coverage at both regional and national scales

[20,21,22].

The dataset includes global horizontal irradiance (GHI), direct normal

irradiance (DNI), diffuse horizontal irradiance (DHI), and beam horizontal

irradiance (BHI), representing the principal components of the surface radiative

balance and forming the basis for solar resource assessment as well as for the design

of photovoltaic and concentrating solar power systems [2,23]. The data are provided

on a regular latitude–longitude grid with a spatial resolution of 0.1° × 0.1°,

corresponding to an average cell size of approximately 10–12 km at the latitude of

Romania, and are available at hourly temporal resolution [20].

For the purposes of this study, the available period 2004–2025 was selected,

providing a 22-year time series suitable for the climatological characterisation of

solar radiation and for the analysis of interannual variability. Hourly values were

extracted for a set of 207 locations distributed across Romania, selected to represent

the diversity of the country’s topographical and climatic conditions, and were

subsequently processed at seasonal and annual scales.

2.2. Time-Series Processing and Variability Indicators

The hourly solar radiation time series were temporally integrated to derive

seasonal and annual totals (kWh/m²), following standard methodologies used in

solar resource assessment. In addition to mean radiation levels, the temporal

stability of the solar resource constitutes an important aspect of energy potential

Geospatial modelling of solar radiation using satellite-derived data

19

analysis. The interannual variability of GHI was assessed by calculating the

standard deviation [24] and the coefficient of variation [25], two statistical

indicators commonly used to characterise the absolute and relative dispersion of

annual values around the multiannual mean.

By combining the multiannual mean with measures of variability, the

analysis identifies areas characterised by high radiation levels and assesses the

degree of interannual stability, which is directly relevant to long-term energy

planning.

2.3. GIS-Based Spatial Processing

The spatial processing and cartographic representation of radiative

parameter distributions were performed in ArcGIS Pro using Romania’s national

Stereo 1970 projection system. Annual and seasonal radiative parameters were

interpolated using Empirical Bayesian Kriging Regression Prediction (EBK-RP), a

geostatistical approach that integrates regression-based modelling with Bayesian

semivariogram estimation [26].

EBK enables the modelling of uncertainty associated with the spatial

autocorrelation structure through the automatic generation of multiple

semivariogram simulations, thereby reducing reliance on fixed assumptions

concerning the spatial distribution of the analysed variable [27]. Recent applications

of EBK-RP in spatial prediction studies indicate that the inclusion of explanatory

covariates may improve interpolation performance relative to EBK, particularly in

contexts characterised by uneven sampling density and by spatial variability

controlled by auxiliary factors [26,28].

The analysis of relationships between radiative variables and geographical

factors was conducted within a GIS framework through a series of spatial

operations and statistical procedures. Point-based data were aggregated into spatial

units (grid cells, elevation classes, and administrative divisions) and subsequently

analysed using zonal statistics. Raster surfaces were compared through cell-by-cell

calculations, raster differencing and ratio analysis, as well as through derived

indicators describing spatial distribution patterns.

Accordingly, the analysis incorporated the Digital Elevation Model (DEM)

together with modelled solar radiation parameters, including Area Solar Radiation

and its direct and diffuse components calculated and applied at seasonal scale, as

explanatory covariates [15], given the influence of topography on radiation

distribution through variations in elevation, slope, and aspect [14]. Final raster

layers were generated at a spatial resolution of 200 m using a mask corresponding

to Romania’s administrative boundaries and uniform processing environment

settings (extent, snap raster, and cell size), thereby ensuring the consistency of the

results.

20

Marina Antonescu, Georgeta Bandoc

The integration of hourly satellite-derived time series, statistical indicators

of variability, and geostatistical modelling within a GIS framework enabled the

spatial and seasonal distribution of the solar resource to be analysed within a unified

analytical framework, suitable for assessment at both national and regional scales.

3. Results and Discussion

3.1. Integrated Spatial Distribution of Radiative Components

Spatial analysis of the radiative components provides insight into the

territorial structure of solar resources and the relationships among the principal

components of the radiative balance. The cartographic representations derived from

the interpolation of multiannual time-series data offers an integrated depiction of

the national-scale distribution of global, direct, and diffuse radiation, thereby

facilitating comparison of their spatial patterns and the identification of areas

characterised by differing radiative potential. These distributions are interpreted

comparatively by examining the absolute values of each component in relation to

the broader spatial structure of the radiative regime.

The multiannual spatial pattern of GHI reveals a marked south-to-north

gradient in solar resources across Romania, together with a pronounced contrast

between the extra-Carpathian regions and the mountainous areas (Fig.1). Maximum

values are concentrated in the southern and south-eastern parts of the country, with

a distinct core over Dobrogea and the Romanian Plain, whereas minimum values

are associated with the Carpathian Mountains and the northern regions. The

observed range (approximately 1098–1502 kWh/m²) indicates substantial spatial

variability of the global solar resource at national scale, reflecting the combined

influence of latitude, regional atmospheric conditions, and orographic factors on

the distribution of solar radiation.

a.

b.

Geospatial modelling of solar radiation using satellite-derived data

21

c.

d.

e.

Fig. 1. Seasonal and annual climatology of multiannual mean global horizontal irradiance (GHI) in

Romania: (a) Winter, (b) Spring, (c) Summer, (d) Autumn, and (e) Annual

(source: created in ArcGIS Pro).

In relation to the overall structure of the global solar resource, the

distribution of the direct component provides a more sensitive perspective on the

influence exerted by atmospheric transparency and cloud regime.

Thus, the annual distribution of DNI follows the same broad spatial

configuration, although territorial contrasts are more pronounced, with high values

concentrated in the southern and south-eastern parts of Romania and a marked

reduction in the mountainous, northern, and north-eastern regions (Fig.2). This

enhancement of spatial contrast is physically consistent, as the direct component is

more sensitive to cloudiness, the frequency of stable synoptic conditions, and

atmospheric transparency. The annual DNI range (approximately 1239–1624

kWh/m²) reflects the marked spatial variability of the direct component and

indicates its substantial contribution to the radiative balance of the southern and

south-eastern regions. At seasonal scale, the same general spatial configuration is

maintained, although the intensity of the contrasts varies according to the radiative

regime specific to each season.

22

Marina Antonescu, Georgeta Bandoc

a.

b.

c.

d.

e.

Fig. 2. Seasonal and annual climatology of multiannual mean direct normal irradiance (DNI) in

Romania: (a) Winter, (b) Spring, (c) Summer, (d) Autumn, and (e) Annual

(source: created in ArcGIS Pro).

Geospatial modelling of solar radiation using satellite-derived data

23

The diffuse component (DHI) exhibits a distinct spatial distribution, with

relatively higher values in the northern and mountainous regions than in southern

Romania (Fig. 3). This configuration reflects the physical complementarity between

the direct and diffuse components of the radiative balance: where DNI is lower, the

relative contribution of diffuse radiation increases, whereas in regions characterised

by higher atmospheric transparency, the direct component contributes more

strongly to the overall radiative balance. In terms of radiative structure, the spatial

relationship among these three components indicates a coherent spatial organisation

of the direct–diffuse ratio and complements the interpretation of the territorial

pattern of GHI.

a.

b.

c.

d.

24

e.

Marina Antonescu, Georgeta Bandoc

Fig. 3. Seasonal and annual climatology of multiannual mean diffuse horizontal irradiance (DHI)

in Romania: (a) Winter, (b) Spring, (c) Summer, (d) Autumn, and (e) Annual

(source: created in ArcGIS Pro).

The spatial distribution of BHI (Fig. 4) is generally aligned with that of DNI,

reflecting the contribution of the direct component to the radiative regime on a

horizontal surface. The highest values are concentrated in the southern and south-

eastern parts of the country, particularly over the Romanian Plain and Dobrogea,

whereas minimum values occur in the north, north-east, and in certain mountainous

areas, where frequent cloudiness and orographic effects reduce the contribution of

direct radiation. The BHI pattern therefore confirms the dominant role of direct

radiation in shaping the high GHI levels observed in the extra-Carpathian regions

and further highlights the spatial consistency of the relationship among the principal

components of the radiative balance.

a.

b.

Geospatial modelling of solar radiation using satellite-derived data

25

c.

d.

e.

Fig. 4. Seasonal and annual climatology of multiannual mean beam horizontal irradiance (BHI) in

Romania: (a) Winter, (b) Spring, (c) Summer, (d) Autumn, and (e) Annual

(source: created in ArcGIS Pro).

Seasonal differentiation of the radiative time series confirms and further

accentuates the spatial contrasts identified at the multiannual scale. During the

warm season, GHI and DNI reach their maximum spatial extent, with high-value

areas expanding across southern and south-eastern Romania, whereas during the

cold season the absolute radiation levels decrease and the influence of regional

atmospheric controls on spatial distribution becomes more pronounced. Under

these conditions, DHI exhibits a relatively more homogeneous distribution in

winter, suggesting an increased contribution of the diffuse component to the

radiative balance under conditions of more frequent cloud cover and reduced

atmospheric transparency.

Considered comparatively, these distributions highlight distinct territorial

differentiations as well as the structural relationships among the components of the

radiative balance. The spatial correspondence between GHI and DNI indicates a

strong positive association, suggesting that areas with high global radiation

potential generally also exhibit high direct radiation potential. However, the

26

Marina Antonescu, Georgeta Bandoc

intensity of the contrasts is greater in the case of DNI, reflecting the higher

sensitivity of this component to atmospheric conditions. In turn, the relationship

between GHI and DHI points to a redistribution of the diffuse fraction in regions

where the direct component is reduced, confirming the structural balance among

the components of the radiative balance. The spatial pattern of BHI further confirms

the role of the projected direct component on the horizontal surface in shaping the

spatial structure of GHI, thereby reinforcing the physical consistency of the

relationships among the analysed variables.

Overall, the results indicate a stable spatial configuration of the solar

resource at the national scale, characterised by a persistent core of favourable

conditions in the southern and south-eastern regions and by relatively low

interannual variability of the global component. The coherent correspondence

among GHI, DNI, DHI, and BHI supports the internal consistency of the dataset

and lends further support to the integrated approach based on satellite-derived data

and geospatial modelling.

3.2. Interannual variability of GHI

The interannual behaviour of GHI, assessed from national mean values

derived from the set of analysed locations, reveals a moderate and statistically

significant positive linear trend. This trend is characterized by a regression slope of

approximately 4.98 kWh/m²/year, a coefficient of determination (R²) of 0.352, and

a significance level of p = 0.00363 (Fig. 5). Statistically, this result indicates that

about 35% of the interannual variability of the mean GHI can be explained by the

linear component of the trend, while the remaining variation reflects year-to-year

fluctuations associated with climate variability.

The annual dispersion, represented by the interval ±1 standard deviation,

indicates moderate interannual fluctuations without evidence of a progressive

increase in spatial heterogeneity. Although certain years (e.g., 2007, 2012, 2022)

show a wider range of variatio, the overall amplitude of the dispersion remains

relatively constant over time, suggesting a stable spatial structure of the resource.

These results indicate that, although mean GHI tends to increase over time, the

degree of spatial differentiation among locations remains relatively stable. From a

methodological perspective, this finding is significant as it indicates that the

increasing signal identified at the mean values is not accompanied by a

corresponding increase in absolute spatial dispersion.

The coefficient of variation (CV) remains within a relatively narrow interval

of approximately 5–7% throughout most of the study period, indicating a moderate

level of relative variability of the GHI at the national scale. The lack of a clear

upward trend in the CV, despite the increase in mean GHI, suggests that the

intensification of radiation level is not accompanied by a proportional increase in

interannual instability.

Geospatial modelling of solar radiation using satellite-derived data

27

Combined analysis of these indicators indicates that the analysed series is

characterised by a gradual increase in mean global radiation, while simultaneously

maintaining a relatively stable structure of spatial dispersion and relative variability.

From an energy-planning perspective, this behaviour may be regarded as favourable

for medium- and long-term decision-making, given the limited degree of

interannual uncertainty.

Fig. 5. Interannual variability and trend of national mean GHI and its spatial variability expressed

in absolute and relative terms.

Conclusions

The results indicate a clearly differentiated spatial distribution of solar

resources across Romania, characterised by a pronounced south-to-north gradient

and persistent contrasts between the extra-Carpathian regions and the mountainous

areas. The high values of GHI, DNI, and BHI in the southern and south-eastern

parts of the country, particularly in Dobrogea and the Romanian Plain, confirm the

favourable solar radiation potential of these areas, whereas the relatively higher

values of DHI in the northern regions and in the mountainous areas reflects the

greater contribution of the diffuse component under less favourable atmospheric

conditions. Comparative analysis of the radiative components further highlights the

internal coherence of the radiation budget and the functional complementarity

between its direct and diffuse components.

At the interannual scale, GHI exhibits a moderate yet statistically significant

positive linear trend, while spatial dispersion and relative variability remain within

moderate bounds. The absence of a systematic increase in the coefficient of

28

Marina Antonescu, Georgeta Bandoc

variation suggests that the intensification of mean global radiation is not

accompanied by a comparable increase in interannual instability.

Overall, the results confirm that the use of multiannual satellite-derived

products, integrated with GIS tools and statistical indicators of variability, provides

a robust methodological framework for solar resource assessment in regions where

dense networks of ground-based radiometric observations are lacking. Beyond the

descriptive value of the maps, this approach enables the identification of structural

relationships among radiative components, the estimate of the interannual stability

of the resource, and the support of further analyses related to energy potential,

project feasibility, and the influence of climatic variability on solar resources.

R E F E R E N C E S

[1]

[2]

[3]

Lewis, N.S., Nocera, D.G., Powering the planet: Chemical challenges in solar energy

utilization, Proc. Natl. Acad. Sci. U.S.A., 103 (43) 15729-15735,

https://doi.org/10.1073/pnas.0603395103 (2006).

Sengupta, M., Habte, A., Gueymard, C., Wilbert, S. Renee, D., Stoffel, T., Best Practices

Handbook for the Collection and Use of Solar Resource Data for Solat Enegy Applications:

Second Edition. NREL (2018).

Dasari, H.P., Desamsetti, S., Langodan, S., Attada, R., Kunchala, R.K., Viswanadhapalli, Y.,

Knio, O., Hoteit, I., High-resolution assessment of solar energy resources over the Arabian

Peninsula, Applied Energy, 248, 354-371, https://doi.org/10.1016/j.apenergy.2019.04.105

(2019).

[4]

Jensen, A.R., Sifnaios, I., Anderson, K.S., Gueymard, C.A., SolarStations.org—A global

catalog of solar irradiance monitoring stations, Solar Energy, 295, 113457,

https://doi.org/10.1016/j.solener.2025.113457 (2025).

[5]

[6]

Bachour, D., Perez-Astudillo, D., Ground-measurement GHI Map for Qatar, Energy

Procedia, 49, 2297-2302, https://doi.org/10.1016/j.egypro.2014.03.243 (2014).

Xu, L., Mao, Y., Evaluation of Two Satellite Surface Solar Radiation Products in the Urban

Region in Beijing, China, Remote Sens., 16(11), 2030, https://doi.org/10.3390/rs16112030

(2024).

[7]

[8]

Sengupta, M., Xie, Y., Lopez, A., Habte, A., Maclaurin, G., Shelby, J., The National Solar

Radiation Data Base (NSRDB), Renewable and Sustainable Energy Reviews, 89, 51-60,

https://doi.org/10.1016/j.rser.2018.03.003 (2018).

Pfeifroth, U., Sanchez-Lorenzo, A., Manara, V., Trentmann, J., Hollmann, R., Trends and

Variability of Surface Solar Radiation in Europe Based On Surface- and Satellite-Based

Data Records, Journal of Geophysical Research: Atmospheres, 123, 1735–1754,

https://doi.org/10.1002/2017JD027418 (2018).

[9]

Lefèvre, M., Oumbe, A., Blanc, P., Espinar, B., Gschwind, B., Qu, Z., Wald, L., Schroedter-

Homscheidt, M., Hoyer-Klick, C., Arola, A., Benedetti, A., Kaiser, J. W., Morcrette, J.-J.,

McClear: a new model estimating downwelling solar radiation at ground level in clear-sky

conditions, Atmos. Meas. Tech., 6, 2403–2418, https://doi.org/10.5194/amt-6-2403-2013

(2013).

[10] Yang, D., Bright, J.M., Worldwide validation of 8 satellite-derived and reanalysis solar

radiation products: A preliminary evaluation and overall metrics for hourly data over 27

years, Solar Energy, 210, 3-19, https://doi.org/10.1016/j.solener.2020.04.016 (2020).

Geospatial modelling of solar radiation using satellite-derived data

29

[11] Gracia Amillo, A.M., Huld, T., Müller, R., A New Database of Global and Direct Solar

Radiation Using the Eastern Meteosat Satellite, Models and Validation, Remote Sens., 6(9),

8165-8189, https://doi.org/10.3390/rs6098165 (2014).

[12] Pfenninger, S., Staffell, I., Long-term patterns of European PV output using 30 years of

validated

hourly

reanalysis

and

satellite

data,

Energy,

114,

1251-1265,

https://doi.org/10.1016/j.energy.2016.08.060 (2016).

[13] Gutiérrez, C., Gaertner, M.Á., Perpiñán, O., Gallardo, C., Sánchez, E., A multi-step scheme

for spatial analysis of solar and photovoltaic production variability and complementarity,

Solar Energy, 158, 100-116, https://doi.org/10.1016/j.solener.2017.09.037 (2017).

[14] Hofierka, J., Šúri, M., The solar radiation model for Open source GIS: implementation and

applications, Proceedings of the Open source GIS - GRASS users conference 2002 - Trento,

Italy (2002).

[15] Anselmo, S., Safaeianpour, A., Moghadam, S.T., Ferrara, M., GIS-based solar radiation

modelling for photovoltaic potential in cities: A sensitivity analysis for the evaluation of

output

https://doi.org/10.1016/j.egyr.2024.10.031 (2024).

[16] Šúri, M., Hofierka, J., A New GIS-based Solar Radiation Model and Its Application to

Photovoltaic Assessments, Transactions in GIS, 8(2):175-190,

https://doi.org/10.1111/j.1467-9671.2004.00174.x (2004).

variability

range,

Energy

Reports,

12,

4656-4669,

[17] Prăvălie, R., Patriche, C., Bandoc, G., Spatial assessment of solar energy potential at global

scale. A geographical approach, Journal of Cleaner Production, 209, 692-721,

https://doi.org/10.1016/j.jclepro.2018.10.239 (2019).

[18] Choi, Y., Suh, J., Kim, S.-M., GIS-Based Solar Radiation Mapping, Site Evaluation, and

Potential Assessment: A Review, Appl. Sci., 9(9), 1960, https://doi.org/10.3390/app9091960

(2019).

[19] Bryce, R., Losada Carreño, I., Kumler, A., Hodge, B.-M., Roberts, B., Martinez-Anido, C.B.,

Consequences of neglecting the interannual variability of the solar resource: A case study of

photovoltaic power among the Hawaiian Islands, Solar Energy, 167, 61-75,

https://doi.org/10.1016/j.solener.2018.03.085 (2018).

[20] Copernicus Atmosphere Monitoring Service (2020): CAMS solar radiation time-series.

Copernicus Atmosphere Monitoring Service (CAMS) Atmosphere Data Store, DOI:

10.24381/5cab0912 (accessed on January 31, 2026).

[21] Schroedter-Homscheidt, M., Azam, F., Betcke, J., Hanrieder, N., Lefèvre, M., Saboret, L.,

Saint-Drenan, Y.-M., Surface solar irradiation retrieval from MSG/SEVIRI based on

APOLLO Next Generation and HELIOSAT-4 methods, Contrib. Atm. Sci./Meteorol. Z.,

doi:10.1127/metz/2022/1132 (2022).

[22] Qu, Z., Oumbe, A., Blanc, P., Espinar, B., Gesell, G., Gschwind, B., Klüser, L., Lefèvre, M.,

Saboret, L., Schroedter-Homscheidt, M., and Wald L., Fast radiative transfer

parameterisation for assessing the surface solar irradiance: The Heliosat-4 method,

Meteorol. Z., 26, 33-57, doi: 10.1127/metz/2016/0781, Available for download at

https://www.schweizerbart.de/papers/metz/detail/26/87036/Fast_radiative_transfer_paramet

erisation_for_asses?af=crossref (2017).

[23] Duffie, J.A., Beckman, W.A., Solar Engineering of Thermal Processes. Chapter 1 - Solar

Radiation, 3-42, John Wiley & Sons, Inc., https://doi.org/10.1002/9781118671603.ch1

(2013).

[24] Standard Deviation. In: The Concise Encyclopedia of Statistics. Springer, New York, NY,

https://doi.org/10.1007/978-0-387-32833-1_381 (2008).

[25] Coefficient of Variation. In: The Concise Encyclopedia of Statistics. Springer, New York,

NY, https://doi.org/10.1007/978-0-387-32833-1_65 (2008).

30

Marina Antonescu, Georgeta Bandoc

[26] ESRI, What is EBK Regression Prediction?, ArcGIS Pro Documentation, Available at:

https://pro.arcgis.com/en/pro-app/latest/help/analysis/geostatistical-analyst/what-is-ebk-

regression-prediction-.htm (accessed on January 31, 2026).

[27] Gribov, A., Krivoruchko, K., Empirical Bayesian kriging implementation and usage, Science

of The Total Environment, 722, 137290, https://doi.org/10.1016/j.scitotenv.2020.137290

(2012).

[28] Njoku, E.A., Akpan, P.E., Effiong, A.E., Babatunde, I.O., The effects of station density in

geostatistical prediction of air temperatures in Sweden: A comparison of two interpolation

techniques,

Resources,

Environment

and

Sustainability,

11,

100092,

DOI:

10.1016/j.resenv.2022.100092 (2023).