Cloud – Index Modelling of Solar Radiation on the Earth’s Surface from Satellite Images.
No Thumbnail Available
Date
2014
Authors
Erusiafe, N.E
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
This work is focused on the extraction of ground-reaching solar radiation from satellite images. The
present methods center mainly on the clear-sky scenario; this work addresses this by developing an all sky
approach. The methodology for extracting ground-reaching solar radiation from satellite images has been
conceptualized such that a cloud index based model was developed for a two-stage process to address the
all-sky condition. The first stage involved characterizing the atmosphere with respect to its solar radiation
transmitting property. Solar radiation measurements were taken at different locations in Nigeria during
various seasons of the year for this purpose. These measurements make up the data from which ground
cloudiness indices were calculated. The characteristic curve, with respect to extra-terrestrial radiation, for
the atmosphere was obtained by plotting loge(G/GO) against the calculated cloudiness index where G and GO are measured ground-reaching solar radiation and extraterrestrial radiation respectively. The groundreaching
radiation may be global solar radiation or spectral components of solar radiation in the visible
spectrum or visible / near infra-red spectrum. The clear sky transmittance was applied in the formulation
of a clear sky model. The clear sky model was applied in characterising the atmosphere by plotting
loge(G/GC) against the calculated cloud index The next stage of the process involved analysing the
reflectance for some locations in Nigeria in level 1.5 images of the Earth and calculating the satellite
cloud index from each image. The reflection in each pixel of the images is presented as digital count and
percentage albedo. The cloud index was determined from the observed reflectance and the reflectances
for the reference clear and cloudy atmospheres. The reflectance for the reference clear atmosphere was
determined from the digital count while that for cloudy atmosphere was taken as the highest possible
reflectance that the satellite’s sensor can register for a given elevation of the Sun. The atmosphere was
also characterized with respect to the derived satellite cloud index. This model was compared with the modified Kasten clear sky model to enable the correlation of the cloud index derived from satellite images
with the attenuating term in the Kasten clear-sky model. This was applied to satellite images to estimate
Linke turbidity factor, TL for Akoka/Lagos. The results obtained from the method in this work show that
the normalized root mean square error (NRMSE) and normalized mean bias error (NMBE) ranges from
0.0667 to 1.4546 and 0.0104 to 0.1819 respectively. The NMBE and NRMSE of the spectral components,
visible/near infra-red (V/NIR) and photosynthetically active radiation (PAR), are 0.0620, 0.0812 and -
0.0750, 0.2568 respectively. The NMBE and NRMSE of the estimated values for Linke turbidity factor
are 0.0013 and 0.2453 respectively. The results obtained from our method compare favourably with results obtained from Solar Data based on Heliosat method – HelioClim. The originality of this work lies
in the fact that it presents a novel approach of exploiting satellite images in the estimation groundreaching
solar radiation and the extraction of information about atmospheric turbidity. An empirically derived clear-sky model that agrees with observed clear sky profile is an outcome of the method proposed
in this work.
Description
A Thesis Submitted to the School of Postgraduate Studies, University of Lagos.
Keywords
Cloud – Index Modelling , Solar Radiation , Satellite Images , Research Subject Categories::NATURAL SCIENCES::Physics
Citation
Erusiafe, N.E (2014), Cloud – Index Modelling of Solar Radiation on the Earth’s Surface from Satellite Images. A Thesis Submitted to University of Lagos School of Postgraduate Studies Phd Thesis and Dissertation, 188pp.