Details

Autor(en) / Beteiligte

• Jin, Yan
• Ge, Yong
• Fan, Haoyu
• Li, Zeshuo
• Jia, Yan
• Liu, Yaojie
• Liu, Hongyan

Titel

Downscaling Solar-Induced Chlorophyll Fluorescence to a 0.05° Monthly Product Using AVHRR Data in East Asia (1995-2003)

Ist Teil von

• IEEE journal of selected topics in applied earth observations and remote sensing, 2024, Vol.17, p.6379-6393

Ort / Verlag

Piscataway: IEEE

Erscheinungsjahr

2024

Quelle

Alma/SFX Local Collection

Beschreibungen/Notizen

• Satellite-based solar-induced chlorophyll fluorescence (SIF) has emerged as a valuable tool for monitoring the photosynthetic activity of vegetation at both regional and global scales. Downscaling techniques offer the opportunity to utilize coarse-spatial-resolution SIF products for investigating carbon cycles and ecological processes at finer resolutions. However, the lack of pre-2000 downscaled products and the limited utilization of residual information in existing models pose challenges in fully harnessing the potential of SIF downscaling. In this study, we generated a new monthly SIF product, DSIF RFK* _EA0.05, at a resolution of 0.05° in East Asia from July 1995 to June 2003. Random forest kriging (RFK) was employed, incorporating GOME SIF, AVHRR data, ERA5 climate data, and using the optimal explanatory variables. Compared with other downscaled results, the produced DSIF RFK* _EA0.05 has higher accuracy and more accurate details of SIF distribution with the highest mean R 2 value of 0.83 and smallest root mean squared error value of 0.08 mW/m 2 /nm/sr. The DSIF RFK* _EA0.05 estimates showed a strong correlation with gross primary productivity data from eight flux sites (R 2 = 0.73), as well as high correlation coefficient values of 0.73, 0.88, and 0.89 with three other fine-resolution products. This study addresses the gap in statistical downscaling of SIF before 2000 and demonstrates the feasibility of utilizing AVHRR data for fine-resolution SIF prediction. Our developed product offers improved detail compared to the original GOME SIF, thereby enhancing the application of satellite SIF for understanding early carbon cycles and ecological processes at a finer resolution.