<trans-title>Retrieval of daily net surface shortwave radiation climatic resources based on machine learning</trans-title>

Wangmin YING; Xiaojie LIU; Shifeng FANG; Xiujuan LI; Ming LAI; Xuzhen ZHANG; Hua WU

doi:10.18402/resci.2020.10.16

Resources Science, Volume. 42, Issue 10, 1998(2020)

Retrieval of daily net surface shortwave radiation climatic resources based on machine learning

Wangmin YING¹, Xiaojie LIU¹, Shifeng FANG¹, Xiujuan LI¹, Ming LAI², Xuzhen ZHANG³, and Hua WU^1,4、*

Author Affiliations

¹Institute of Geographic Sciences and Natural Resources Research, CAS, Beijing 100101, China

²School of Earth Resources, China University of Geosciences (Wuhan), Wuhan 430000, China

³Yantai Coastal Zone Geological Survey Center, China Geological Survey, Yantai 264004, China

⁴College of Resources and Environment, University of Chinese Academy of Sciences, Beijing 100049, China

show less

Abstract Get PDF(in Chinese)

Figures & Tables(9)

Fig. 1. Distribution of the 95 selected sites from the FLUXNET 2015 dataset and the International Geosphere-Biosphere Programme (IGBP) surface types

Download full size

View in Article

Fig. 2. Flowchart for the development and validation of the daily net surface shortwave radiation (DNSSR) model

Download full size

View in Article

Fig. 3. Comparison of daily net surface shortwave radiation (DNSSR) values estimated using the random forest (RF) method with in-situ NSSRs

Download full size

View in Article

Fig. 4. Distribution of global daily net surface shortwave radiation (DNSSR) retrieval

Download full size

View in Article

Fig. 5. Distribution of global daily net surface shortwave radiation (DNSSR) from ERA5 data

Download full size

View in Article

Fig. 6. Comparison of daily net surface shortwave radiation (DNSSR) from ERA5 data and FLUXNET observations

Download full size

View in Article

Table 1. Statistics of daily net surface shortwave radiation retrieval accuracy

View table

View in Article

Table 1. Statistics of daily net surface shortwave radiation retrieval accuracy

传感器	分辨率	重返周期/天	RMSE/（W/m²）	站点数&时间跨度	方法类型	参考文献
MODIS/ Terra	1 km	1	40.5（全天候）	7个站点&1年	混合型	Kim et al^[8]
Landsat 7	30 m	16	36.1（全天候）	6个站点&1年	混合型	Wang et al^[15]
MODIS/ Combined	1 km	1	42.6（有云）	7个站点&1年	混合型	Tang et al^[18]
FY3B/ MERSI	1 km	1	41.9（全天候）	7个站点&1年	混合型	Wang et al^[16]
MODIS/ Combined	1 km	1	28.7（全天候）	7个站点&1年	混合型	Wang et al^[17]
Landsat 7	30 m	16	18.8（全天候）	171个站点&数年	机器学习	Wang et al^[19]

Table 2. Statistics of surface type of selected FLUXNET 95 sites
View table
View in Article
Table 2. Statistics of surface type of selected FLUXNET 95 sites
地表类型 ENF OSH EBF WSA DNF SAV DBF GRA MF WET CSH CRO
站点数量 19 4 9 4 1 4 11 14 6 12 1 10

Table 3. Error statistics of ERA5-daily net surface shortwave radiation (DNSSR) correlation with MODIS-DNSSR
View table
View in Article
Table 3. Error statistics of ERA5-daily net surface shortwave radiation (DNSSR) correlation with MODIS-DNSSR
20190101 20190401 20190701 20191001
相关系数r 0.90 0.94 0.82 0.94
偏差Bias/(W/m²) 1.2 -13.3 10.1 -24.0
均方根误差RMSE/ (W/m²) 32.2 31.9 40.6 36.9
决定系数R² 0.78 0.83 0.51 0.74

Tools

Get Citation

Copy Citation Text

Wangmin YING, Xiaojie LIU, Shifeng FANG, Xiujuan LI, Ming LAI, Xuzhen ZHANG, Hua WU. Retrieval of daily net surface shortwave radiation climatic resources based on machine learning[J]. Resources Science, 2020, 42(10): 1998

Download Citation

EndNote(RIS)BibTex Plain Text

Set citation alerts for article

Save article for my favorites