Long-term effects of rewetting and drought on GPP in a temperate peatland based on satellite remote sensing data

Yinying Zhou; Torsten Sachs; Zhan Li; Yuwen Pang; Junfeng Xu; Aram Kalhori; Christian Wille; Xiaoxue Peng; Xianhao Fu; Yanfei Wu; Lin Wu

doi:10.1016/j.scitotenv.2023.163395

Long-term effects of rewetting and drought on GPP in a temperate peatland based on satellite remote sensing data

Sci Total Environ. 2023 Jul 15:882:163395. doi: 10.1016/j.scitotenv.2023.163395. Epub 2023 Apr 11.

Authors

Yinying Zhou¹, Torsten Sachs², Zhan Li², Yuwen Pang³, Junfeng Xu⁴, Aram Kalhori², Christian Wille², Xiaoxue Peng⁵, Xianhao Fu⁵, Yanfei Wu⁵, Lin Wu⁶

Affiliations

¹ School of information science and technology, Hangzhou Normal University, Hangzhou 311121, China; Ningbo Alatu Digital Technology Co., Ltd., Ningbo, China.
² GFZ German Research Centre for Geosciences, Potsdam, Germany.
³ Ecosystems and Environment Research Programme, Faculty of Biological and Environmental Sciences, Helsinki Institute of Sustainability Science (HELSUS), University of Helsinki, Helsinki, Finland.
⁴ School of information science and technology, Hangzhou Normal University, Hangzhou 311121, China. Electronic address: xjf11@zju.edu.cn.
⁵ School of information science and technology, Hangzhou Normal University, Hangzhou 311121, China.
⁶ Hubei Minzu University, Enshi, China.

PMID: 37044335
DOI: 10.1016/j.scitotenv.2023.163395

Abstract

Rewetting previously drained peatlands restores the critical function of peatlands as long-term carbon storages and sinks currently threatened by climate change and additional human-induced disturbances. Understanding and projecting the restoration process by rewetting, however, currently face a pressing challenge, the lack of consistent and gap-free records of important carbon cycling indicators of peatlands such as the gross primary production (GPP) over long term. In this study, we reconstructed the GPP in a rewetted peatland called Zarnekow (Fluxnet-ID: DE-Zrk) in Germany from 2000 to 2020 by combining long-term satellite observations and limited-term tower-based eddy covariance (EC) measurements based on Random Forest regression models. The R² between the reconstructed data and EC data was 0.6. The reasonable reconstruction of long-term GPP enabled trend analysis that identified two distinct periods of decreasing/increasing in GPP due to rewetting and droughts. Rewetting in the winter of 2004 and 2005 stabilized GPP after a decreasing period. A drought in 2018 significantly increased GPP, and GPP remained high over the following two years. Furthermore, the month-specific trends show significant seasonality at this site, specifically, an increasing trend over the 21 years in the growing-season months of June to August and a decreasing trend in the other months. The most important variables for satellite-based estimates of GPP at this site include total evapotranspiration, land surface temperature, enhanced vegetation index and near-infrared reflectance vegetation index. Long-term analyses of carbon fluxes through the combination of satellite observations and EC measurements provide crucial insights into the restoration of carbon sequestration functions in rewetted peatlands.

Keywords: Data reconstruction; Gross primary production; Peatland; Random Forest; Remote sensing retrieval.