Recent Amplified Global Gross Primary Productivity Due to Temperature Increase Is Offset by Reduced Productivity Due to Water Constraints

Abstract

Satellite remote sensing observations show an increased greenness trend over land in recent decades. While greenness observations can indicate increased productivity, estimation of total annual productivity is highly dependent on vegetation response to climate and environmental conditions. Models have been struggling to determine how much carbon is taken up by plants as a result of increased atmospheric CO 2 fertilization. Current remote sensing light use efficiency (LUE) models contain considerable uncertainty due to the lack of spatial and temporal variability in maximum LUE parameter and climate sensitivity defined for global plant functional types (PFTs). We used the optimum LUE (LUE opt) previously derived from the global FLUXNET network to improve estimation of global gross primary productivity (GPP) for the period 1982-2016. Our results indicate increasing GPP in northern latitudes owing to reduced cold temperature constraints on plant growth, thereby suggesting increasing negative carbon-climate feedback in high latitudes. In the tropics, by contrast, our results indicate an emerging positive climate feedback, mainly due to increasing atmospheric vapor pressure deficit (VPD). Further pervasive VPD increase is likely to continue to reduce global GPP and amplify carbon emissions. Plain Language Summary In light use efficiency (LUE) models, plant production is linearly related to canopy absorbed photosynthetically active radiation (APAR), based on the assumption that plants absorb and convert solar radiant energy into vegetation biomass with a given efficiency rate. Here, we used an enhanced LUE model driven with remote sensing observations to estimate plant productivity for 1982-2016. We found that over the study period, plant photosynthetic activity has increased over northern latitudes, which may partially offset the CO 2 emissions from fossil fuel consumption. However, our results show that productivity in the tropical zones is declining rapidly due to increased water stress. With increased warming, water limitations are expected to increasingly limit global plant productivity.

Divergent historical GPP trends among state-of-the-art multi-model simulations and satellite-based products

“…S4, Fig. S6) (Hashimoto et al, 2013;Yuan et al, 2019;Madani et al, 2020). Between 2001 and 2015, the DGVM ensemble mean GPP trend increases more than NIRv over tropical regions and is consistent with NIRv in middle and high latitudes.…”

“…Although there have been a lot of studies focusing on extreme anomalies, the seasonal cycle, interannual variation, and the climatological pattern of global and regional GPP based on the multiple GPP products and proxy indices (Chen et al, 2017;Madani et al, 2020;Wang et al, 2021b), few efforts have been devoted to evaluate the long-term GPP trends across different GPP sources and to analyze the causes of uncertainties. This study comprehensively investigates historical GPP trends during 1982−2015, based on the satellite-derived GPP proxy (NIRv), TRENDYv6 multi-model simulations, machine-learning products, satellite-based estimates, and site-level observations.…”

Detecting vulnerability of humid tropical forests to multiple stressors

“…For forest GPP, we used recently developed enhanced global GPP from a combination of remote-sensing, climate, and eddy covariance tower observations globally. 78 The enhanced GPP data product is developed from Global Inventory Modeling and Mapping Studies (GIMMS3g) fraction of photosynthetically active radiation record for the period 1982-2016 79 extended by the MODIS data to 2018, and solar-induced chlorophyll fluorescence from the ll OPEN…”

“…Global Ozone Monitoring Experiment-2. 80 An optimum light use efficiency (LUE) model previously derived from the global FLUXNET networks 78,81 was used as the primary model to integrate remote-sensing and climate data (VPD, radiation, soil moisture) to estimate long-term (1982-2018) GPP globally at 8-km spatial and biweekly temporal resolutions.…”

“…Building on the assumption that we have zero covariance terms for both residual errors between products, and between errors and truth, the pixel level error for each product can be expressed using sample covariance matrices between participating products. We therefore estimated the GPP uncertainty of the enhanced GPP data product 78 using two other independent GPP products, GOSIF 94 and FluxCom, 95 for the overlapped period (2002)(2003)(2004)(2005)(2006)(2007)(2008)(2009)(2010)(2011)(2012)(2013)(2014)(2015) to form the triplets. To estimate the uncertainty of ET, we collected additional ET products from ERA5 and MODIS and used the GLEAM ET from our study to perform triple collocation for the overlapped period (2002)(2003)(2004)(2005)(2006)(2007)(2008)(2009)(2010)(2011)(2012)(2013)(2014)(2015)(2016)(2017)(2018).…”