This article has been accepted for inclusion in a future issue of this journal. Content is final as presented, with the exception of pagination.
IEEE GEOSCIENCE AND REMOTE SENSING LETTERS 1
Impacts of DEM Geolocation Bias on Downward
Surface Shortwave Radiation Estimation Over
Clear-Sky Rugged Terrain: A Case Study
in Dayekou Basin, China
Dalei Hao , Jianguang Wen , Qing Xiao , Shengbiao Wu , Xingwen Lin, Dongqin You, and Yong Tang
Abstract— Accurately estimating the spatial–temporal dis-
tribution of downward surface shortwave radiation (DSSR)
is essential for terrestrial ecological modeling and climate
change research. The accurate georegistration of digital elevation
model (DEM) has become one of the significant bottlenecks for
improving the DSSR accuracy over rugged terrain. To clearly
understand and quantitatively evaluate the impact of geolocation
bias on the DSSR estimation under clear sky, this letter conducts
a systematical simulation research in Dayekou Basin of China
based on a developed remote sensing satellite-based DSSR esti-
mation scheme over rugged terrain. The results demonstrate that
the proposed approach can accurately capture the high temporal
and spatial heterogeneities of DSSR, and the DSSR estimations
are sensitive to geolocation bias. When the horizontal bias is lower
than half a pixel, the deviations of the direct radiation could lead
to above 600 W/m
2
due to the illumination angle effects and
shadow effects. The consequence of the bias on the diffuse and
reflected radiation from adjacent terrains is little because of their
relatively small values and low-spatial heterogeneities under clear
sky in general except for the deep valley areas. The trends of the
total radiation errors with the geolocation bias are identical in
different days (scenes), and the error is related to the solar zenith
angle. In addition, the more rugged the terrain, the greater the
influence of geolocation bias on the radiation accuracy.
Index Terms—Downward surface shortwave radiation (DSSR),
geolocation bias, rugged terrain, topographic effect.
Manuscript received April 20, 2018; revised August 1, 2018; accepted
August 31, 2018. This work was supported by the Chinese Natural Sci-
ence Foundation Project under Grant 41671363. (Corresponding author:
Jianguang Wen.)
D. Hao, Q. Xiao, S. Wu, and X. Lin are with the State Key Laboratory
of Remote Sensing Science, Institute of Remote Sensing and Digital Earth,
Chinese Academy of Sciences, Beijing 100101, China, and also with the
College of Resources and Environment, University of Chinese Academy of
Sciences, Beijing 100049, China.
J. Wen is with the State Key Laboratory of Remote Sensing Science,
Institute of Remote Sensing and Digital Earth, Chinese Academy of Sciences,
Beijing 100101, China, with the College of Resources and Environment,
University of Chinese Academy of Sciences, Beijing 100049, China, and also
with the Joint Center for Global Change Studies, Beijing 100875, China.
(e-mail: wenjg@radi.ac.cn).
D. You and Y. Tang are with the State Key Laboratory of Remote Sensing
Science, Institute of Remote Sensing and Digital Earth, Chinese Academy of
Sciences, Beijing 100101, China.
Color versions of one or more of the figures in this letter are available
online at http://ieeexplore.ieee.org.
Digital Object Identifier 10.1109/LGRS.2018.2868563
I. I NTRODUCTION
D
OWNWARD surface shortwave radiation (DSSR) is a
critical variable in global change sciences, which signif-
icantly influences the earth’s energy exchange [1]. Accurate
characterization of the spatial and temporal distribution of
DSSR over the globe is the fundamental and essential for
meteorology and oceanography, as well as for agricultural,
social applications, and fisheries [2]. Remote sensing provides
a unique advantage in mapping regional and global DSSRs at
continuous spatial and temporal scales.
Over the years, many space-based algorithms for estimating
DSSR over flat terrain have been developed from various
remote sensing observations [3]–[5]. Compared to flat sur-
faces, the interaction between shortwave radiation and the
earth’s surface is particularly complex over mountainous areas.
Neglecting the topographic effects can lead to considerable
uncertainty and significant errors in the derived DSSR over
rugged terrain [6], [7]. Accurate estimation of DSSR over
rugged terrain presents a challenge for remote sensing com-
munities. Recently, the availability of high-quality digital
elevation model (DEM) with high spatial resolution promotes
the development of DSSR estimation algorithms over rugged
terrain [8]. However, it is worth noting that the uncertainty
of DEM data has great impacts on the estimation accuracy
of DSSR. Different DEM data sets with identical absolute
altitudes could lead to significant differences in the derived
DSSRs [9] and DEM resolution also plays an important
role in DSSR determination [10]. In addition, the accurate
georegistration of DEM is an essential precursor to the DSSR
estimation over rugged terrain. However, the quantitative con-
tributions of the DEM geolocation bias on the quality and
accuracy of DSSR estimation remain unknown. This letter
aims to quantitatively understand and evaluate the conse-
quences of DEM geolocation bias on DSSR estimation over
rugged terrain.
II. METHODS
A. DEM Geolocation Bias Simulation
To simulate the DEM geolocation bias, the DEM is first
georegistered by using ground control points. Assuming that
the geolocation noise is uniformly distributed spatially in
a regional neighborhood, only local translation geolocation
1545-598X © 2018 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission.
See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.