A novel approach for ecosystem respiration simulation in drylands

• DOI: 10.3389/fevo.2023.1186272
• Publication Date: 2023-10-03T04:36:31Z
• Authors: Jinlong Ai; Xiaowen Qi; Rensen Zhang; Mingye He; Jingyang Li; Ronghan Xu; Yapeng Li; Sangeeta Sarmah; Huan Wang; Junfang Zhao
• Keywords: 0301 basic medicine; Atmospheric sciences; Ecosystem Resilience; Evolution; Arid; FLUXNET; Climate Change and Variability Research; Environmental science; remote sensing; 03 medical and health sciences; Biome; QH359-425; Biology; QH540-549.5; Ecosystem; Climatology; drylands; Global and Planetary Change; Ecology; Geography; Global Forest Drought Response and Climate Change; Physics; Geology; Carbon cycle; FOS: Earth and related environmental sciences; Remote sensing; terrestrial ecosystem respiration; simulation; FOS: Biological sciences; Environmental Science; Physical Sciences; Global Methane Emissions and Impacts; Climate Modeling

Terrestrial ecosystem respiration (Reco) in drylands (arid and semi-arid areas) contributes to the largest uncertainty of the global carbon cycle. Here, using the Reco data from 24 sites (98 site-years) in drylands from Fluxnet and corresponding MODIS remote sensing products, we develop a novel semi-empirical, yet physiologically-based remote sensing model: the ILEP_Reco model (a Reco model derived from ILEP, the acronym for “integrated LE and EVI proxy”). This model can simulate Reco observations across most biomes in drylands with a small margin of error (R2 = 0.56, RMSE = 1.12 gCm−2d−1, EF = 0.46, MBE = −0.06 gCm−2d−1) and performs significantly better than the previous model: Ensemble_all. The seasonal variation of Reco in drylands can be well simulated by the ILEP_Reco model. When we relate ILEP to the Q10 model, the corresponding ILEP_Q10 values in all 98 site-years distribute quite convergently, which greatly facilitates fixing the ILEP_Q10 value as a constant in different site-years. The spatial variation of Reco in drylands is then defined as reference respiration at the annual mean ILEP, which can be easily and powerfully simulated by the ILEP_Reco model. These results help us understand the spatial-temporal variations of Reco in drylands and thus will shed light on the carbon budget on a regional scale, or even a global one.