A global long-term (1981–2019) daily land surface radiation budget product from AVHRR satellite data using a residual convolutional neural network
<p>The surface radiation budget, also known as all-wave net radiation (<span class="inline-formula"><i>R</i><sub>n</sub></span>), is a key parameter for various land surface processes including hydrological, ecological, agricultural, and biogeochem...
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Copernicus Publications
2022-05-01
|
| Series: | Earth System Science Data |
| Online Access: | https://essd.copernicus.org/articles/14/2315/2022/essd-14-2315-2022.pdf |
| _version_ | 1811258308567236608 |
|---|---|
| author | J. Xu S. Liang B. Jiang |
| author_facet | J. Xu S. Liang B. Jiang |
| author_sort | J. Xu |
| collection | DOAJ |
| description | <p>The surface radiation budget, also known as
all-wave net radiation (<span class="inline-formula"><i>R</i><sub>n</sub></span>), is a key parameter for various land
surface processes including hydrological, ecological, agricultural, and
biogeochemical processes. Satellite data can be effectively used to estimate
<span class="inline-formula"><i>R</i><sub>n</sub></span>, but existing satellite products have coarse spatial resolutions and
limited temporal coverage. In this study, a point-surface matching
estimation (PSME) method is proposed to estimate surface <span class="inline-formula"><i>R</i><sub>n</sub></span> using a
residual convolutional neural network (RCNN) integrating spatially adjacent
information to improve the accuracy of retrievals. A global high-resolution
(0.05<span class="inline-formula"><sup>∘</sup></span>), long-term (1981–2019), and daily mean <span class="inline-formula"><i>R</i><sub>n</sub></span> product
was subsequently generated from Advanced Very High Resolution Radiometer
(AVHRR) data. Specifically, the RCNN was employed to establish a nonlinear
relationship between globally distributed ground measurements from 522 sites
and AVHRR top-of-atmosphere (TOA) observations. Extended triplet collocation
(ETC) technology was applied to address the spatial-scale mismatch issue
resulting from the low spatial support of ground measurements within the
AVHRR footprint by selecting reliable sites for model training. The overall
independent validation results show that the generated AVHRR <span class="inline-formula"><i>R</i><sub>n</sub></span> product
is highly accurate, with <span class="inline-formula"><i>R</i><sup>2</sup></span>, root-mean-square error (RMSE), and bias of
0.84, 26.77 W m<span class="inline-formula"><sup>−2</sup></span> (31.54 %), and 1.16 W m<span class="inline-formula"><sup>−2</sup></span> (1.37 %),
respectively. Inter-comparisons with three other <span class="inline-formula"><i>R</i><sub>n</sub></span> products, i.e., the
5 km Global Land Surface Satellite (GLASS); the 1<span class="inline-formula"><sup>∘</sup></span> Clouds and the
Earth's Radiant Energy System (CERES); and the 0.5<span class="inline-formula"><sup>∘</sup></span> <span class="inline-formula">×</span> 0.625<span class="inline-formula"><sup>∘</sup></span> Modern-Era Retrospective analysis for Research and
Applications, Version 2 (MERRA-2), illustrate that our AVHRR <span class="inline-formula"><i>R</i><sub>n</sub></span>
retrievals have the best accuracy under most of the considered surface and
atmospheric conditions, especially thick-cloud or hazy conditions. However,
the performance of the model needs to be further improved for the snow/ice
cover surface. The spatiotemporal analyses of these four <span class="inline-formula"><i>R</i><sub>n</sub></span> datasets
indicate that the AVHRR <span class="inline-formula"><i>R</i><sub>n</sub></span> product reasonably replicates the spatial
pattern and temporal evolution trends of <span class="inline-formula"><i>R</i><sub>n</sub></span> observations. The long-term
record (1981–2019) of the AVHRR <span class="inline-formula"><i>R</i><sub>n</sub></span> product shows its value in climate
change studies. This dataset is freely available at
<a href="https://doi.org/10.5281/zenodo.5546316">https://doi.org/10.5281/zenodo.5546316</a> for 1981–2019 (Xu et al.,
2021).</p> |
| first_indexed | 2024-04-12T18:12:21Z |
| format | Article |
| id | doaj.art-9362ae834c2a4bdbb9324abe3b20e86d |
| institution | Directory Open Access Journal |
| issn | 1866-3508 1866-3516 |
| language | English |
| last_indexed | 2024-04-12T18:12:21Z |
| publishDate | 2022-05-01 |
| publisher | Copernicus Publications |
| record_format | Article |
| series | Earth System Science Data |
| spelling | doaj.art-9362ae834c2a4bdbb9324abe3b20e86d2022-12-22T03:21:48ZengCopernicus PublicationsEarth System Science Data1866-35081866-35162022-05-01142315234110.5194/essd-14-2315-2022A global long-term (1981–2019) daily land surface radiation budget product from AVHRR satellite data using a residual convolutional neural networkJ. Xu0S. Liang1B. Jiang2School of Remote Sensing and Information Engineering, Wuhan University, Wuhan 430079, ChinaDepartment of Geographical Sciences, University of Maryland, College Park, MD 20742, USAFaculty of Geographical Science, Beijing Normal University, Beijing 100875, China<p>The surface radiation budget, also known as all-wave net radiation (<span class="inline-formula"><i>R</i><sub>n</sub></span>), is a key parameter for various land surface processes including hydrological, ecological, agricultural, and biogeochemical processes. Satellite data can be effectively used to estimate <span class="inline-formula"><i>R</i><sub>n</sub></span>, but existing satellite products have coarse spatial resolutions and limited temporal coverage. In this study, a point-surface matching estimation (PSME) method is proposed to estimate surface <span class="inline-formula"><i>R</i><sub>n</sub></span> using a residual convolutional neural network (RCNN) integrating spatially adjacent information to improve the accuracy of retrievals. A global high-resolution (0.05<span class="inline-formula"><sup>∘</sup></span>), long-term (1981–2019), and daily mean <span class="inline-formula"><i>R</i><sub>n</sub></span> product was subsequently generated from Advanced Very High Resolution Radiometer (AVHRR) data. Specifically, the RCNN was employed to establish a nonlinear relationship between globally distributed ground measurements from 522 sites and AVHRR top-of-atmosphere (TOA) observations. Extended triplet collocation (ETC) technology was applied to address the spatial-scale mismatch issue resulting from the low spatial support of ground measurements within the AVHRR footprint by selecting reliable sites for model training. The overall independent validation results show that the generated AVHRR <span class="inline-formula"><i>R</i><sub>n</sub></span> product is highly accurate, with <span class="inline-formula"><i>R</i><sup>2</sup></span>, root-mean-square error (RMSE), and bias of 0.84, 26.77 W m<span class="inline-formula"><sup>−2</sup></span> (31.54 %), and 1.16 W m<span class="inline-formula"><sup>−2</sup></span> (1.37 %), respectively. Inter-comparisons with three other <span class="inline-formula"><i>R</i><sub>n</sub></span> products, i.e., the 5 km Global Land Surface Satellite (GLASS); the 1<span class="inline-formula"><sup>∘</sup></span> Clouds and the Earth's Radiant Energy System (CERES); and the 0.5<span class="inline-formula"><sup>∘</sup></span> <span class="inline-formula">×</span> 0.625<span class="inline-formula"><sup>∘</sup></span> Modern-Era Retrospective analysis for Research and Applications, Version 2 (MERRA-2), illustrate that our AVHRR <span class="inline-formula"><i>R</i><sub>n</sub></span> retrievals have the best accuracy under most of the considered surface and atmospheric conditions, especially thick-cloud or hazy conditions. However, the performance of the model needs to be further improved for the snow/ice cover surface. The spatiotemporal analyses of these four <span class="inline-formula"><i>R</i><sub>n</sub></span> datasets indicate that the AVHRR <span class="inline-formula"><i>R</i><sub>n</sub></span> product reasonably replicates the spatial pattern and temporal evolution trends of <span class="inline-formula"><i>R</i><sub>n</sub></span> observations. The long-term record (1981–2019) of the AVHRR <span class="inline-formula"><i>R</i><sub>n</sub></span> product shows its value in climate change studies. This dataset is freely available at <a href="https://doi.org/10.5281/zenodo.5546316">https://doi.org/10.5281/zenodo.5546316</a> for 1981–2019 (Xu et al., 2021).</p>https://essd.copernicus.org/articles/14/2315/2022/essd-14-2315-2022.pdf |
| spellingShingle | J. Xu S. Liang B. Jiang A global long-term (1981–2019) daily land surface radiation budget product from AVHRR satellite data using a residual convolutional neural network Earth System Science Data |
| title | A global long-term (1981–2019) daily land surface radiation budget product from AVHRR satellite data using a residual convolutional neural network |
| title_full | A global long-term (1981–2019) daily land surface radiation budget product from AVHRR satellite data using a residual convolutional neural network |
| title_fullStr | A global long-term (1981–2019) daily land surface radiation budget product from AVHRR satellite data using a residual convolutional neural network |
| title_full_unstemmed | A global long-term (1981–2019) daily land surface radiation budget product from AVHRR satellite data using a residual convolutional neural network |
| title_short | A global long-term (1981–2019) daily land surface radiation budget product from AVHRR satellite data using a residual convolutional neural network |
| title_sort | global long term 1981 2019 daily land surface radiation budget product from avhrr satellite data using a residual convolutional neural network |
| url | https://essd.copernicus.org/articles/14/2315/2022/essd-14-2315-2022.pdf |
| work_keys_str_mv | AT jxu agloballongterm19812019dailylandsurfaceradiationbudgetproductfromavhrrsatellitedatausingaresidualconvolutionalneuralnetwork AT sliang agloballongterm19812019dailylandsurfaceradiationbudgetproductfromavhrrsatellitedatausingaresidualconvolutionalneuralnetwork AT bjiang agloballongterm19812019dailylandsurfaceradiationbudgetproductfromavhrrsatellitedatausingaresidualconvolutionalneuralnetwork AT jxu globallongterm19812019dailylandsurfaceradiationbudgetproductfromavhrrsatellitedatausingaresidualconvolutionalneuralnetwork AT sliang globallongterm19812019dailylandsurfaceradiationbudgetproductfromavhrrsatellitedatausingaresidualconvolutionalneuralnetwork AT bjiang globallongterm19812019dailylandsurfaceradiationbudgetproductfromavhrrsatellitedatausingaresidualconvolutionalneuralnetwork |