Performance Analysis of Empirical Models for Daily Global Solar Radiation in Jiri, Nepal

Abstract

Accurate estimation of global solar radiation (GSR) is essential for solar energy system design, climate analysis, and sustainable energy planning. In mountainous countries such as Nepal, direct measurement of solar radiation is limited due to high instrumentation costs and sparse monitoring networks. As a result, empirical models based on routinely measured meteorological parameters provide a practical alternative for solar resource assessment. This study evaluates the performance of twenty empirical models for estimating daily global solar radiation in Jiri, a mid-hill region of eastern Nepal characterized by moderate altitude, frequent cloud cover, and strong monsoonal influence.
Long-term meteorological data, including sunshine duration, air temperature, and relative humidity, were used as model inputs. Extraterrestrial solar radiation and day length were calculated using standard astronomical relations, and model coefficients were derived through regression analysis. Model performance was assessed primarily using error-based statistical indicators, namely Root Mean Square Error (RMSE), Mean Bias Error (MBE), and Mean Percentage Error (MPE) and coefficient of determination (R2),which are more appropriate than correlation-based measures under complex climatic and topographic conditions.
The results show that the Modified Angstrom (new) models M14 and M15 outperform the other models, exhibiting the lowest RMSE values (3.63–3.67 MJ m⁻ ² day⁻ ¹), minimum MPE (approximately 9.2–9.4%), and negligible bias, indicating high accuracy and long-term stability. Models M17, M18, and M20 also demonstrate satisfactory performance and can be considered reliable alternatives. In contrast, simpler sunshine- or temperature-based models show relatively higher errors and reduced suitability for the Jiri region.
The average annual global solar radiation in Jiri was estimated as 15.05 MJ m⁻ ² day⁻ ¹ (approximately 4.18 kWh m⁻ ² day⁻ ¹), confirming significant solar energy potential. The findings highlight the importance of site-specific validation and support the use of multi-parameter empirical models for solar energy planning in Nepal’s mid-hill regions.