Journal cover Journal topic
Atmospheric Chemistry and Physics An interactive open-access journal of the European Geosciences Union
Journal topic

Journal metrics

IF value: 5.414
IF5.414
IF 5-year value: 5.958
IF 5-year
5.958
CiteScore value: 9.7
CiteScore
9.7
SNIP value: 1.517
SNIP1.517
IPP value: 5.61
IPP5.61
SJR value: 2.601
SJR2.601
Scimago H <br class='widget-line-break'>index value: 191
Scimago H
index
191
h5-index value: 89
h5-index89
Volume 16, issue 9
Atmos. Chem. Phys., 16, 5949–5967, 2016
https://doi.org/10.5194/acp-16-5949-2016
© Author(s) 2016. This work is distributed under
the Creative Commons Attribution 3.0 License.
Atmos. Chem. Phys., 16, 5949–5967, 2016
https://doi.org/10.5194/acp-16-5949-2016
© Author(s) 2016. This work is distributed under
the Creative Commons Attribution 3.0 License.

Research article 17 May 2016

Research article | 17 May 2016

Implementation of Bessel's method for solar eclipses prediction in the WRF-ARW model

Alex Montornès1, Bernat Codina1, John W. Zack2, and Yolanda Sola1 Alex Montornès et al.
  • 1Department of Astronomy and Meteorology, University of Barcelona, Barcelona, Spain
  • 2MESO Inc., Troy, USA

Abstract. Solar eclipses are predictable astronomical events that abruptly reduce the incoming solar radiation into the Earth's atmosphere, which frequently results in non-negligible changes in meteorological fields. The meteorological impacts of these events have been analyzed in many studies since the late 1960s. The recent growth in the solar energy industry has greatly increased the interest in providing more detail in the modeling of solar radiation variations in numerical weather prediction (NWP) models for the use in solar resource assessment and forecasting applications. The significant impact of the recent partial and total solar eclipses that occurred in the USA (23 October 2014) and Europe (20 March 2015) on solar power generation have provided additional motivation and interest for including these astronomical events in the current solar parameterizations.

Although some studies added solar eclipse episodes within NWP codes in the 1990s and 2000s, they used eclipse parameterizations designed for a particular case study. In contrast to these earlier implementations, this paper documents a new package for the Weather Research and Forecasting–Advanced Research WRF (WRF-ARW) model that can simulate any partial, total or hybrid solar eclipse for the period 1950 to 2050 and is also extensible to a longer period. The algorithm analytically computes the trajectory of the Moon's shadow and the degree of obscuration of the solar disk at each grid point of the domain based on Bessel's method and the Five Millennium Catalog of Solar Eclipses provided by NASA, with a negligible computational time. Then, the incoming radiation is modified accordingly at each grid point of the domain.

This contribution is divided in three parts. First, the implementation of Bessel's method is validated for solar eclipses in the period 1950–2050, by comparing the shadow trajectory with values provided by NASA. Latitude and longitude are determined with a bias lower than 5  ×  10−3 degrees (i.e.,  ∼  550 m at the Equator) and are slightly overestimated and underestimated, respectively. The second part includes a validation of the simulated global horizontal irradiance (GHI) for four total solar eclipses with measurements from the Baseline Surface Radiation Network (BSRN). The results show an improvement in mean absolute error (MAE) from 77 to 90 % under cloudless skies. Lower agreement between modeled and measured GHI is observed under cloudy conditions because the effect of clouds is not included in the simulations for a better analysis of the eclipse outcomes. Finally, an introductory discussion of eclipse-induced perturbations in the surface meteorological fields (e.g., temperature, wind speed) is provided by comparing the WRF–eclipse outcomes with control simulations.

Publications Copernicus
Download
Short summary
This paper documents a new package for the Weather Research and Forecasting--Advanced Research WRF (WRF-ARW) model that can simulate any partial, total or hybrid solar eclipse for the period 1950–2050 and is also extensible to a longer period. First, a description of the implementation together with a validation for the period 1950–2050 of all solar eclipse trajectories is presented. Second, the model response is analyzed in four total solar eclipse episodes. Global horizontal irradiance (GHI) outcomes are validated with respect to ground-based measurements.
This paper documents a new package for the Weather Research and Forecasting--Advanced Research...
Citation
Altmetrics
Final-revised paper
Preprint