Seasonal Responses of Terrestrial Carbon Cycle to Climate Variations in CMIP5 Models: Evaluation and Projection

Abstract

Seventeen Earth system models (ESMs) from phase 5 of the Coupled Model Intercomparison Project (CMIP5) were evaluated, focusing on the seasonal sensitivities of net biome production (NBP), net primary production (NPP), and heterotrophic respiration (Rh) to interannual variations in temperature and precipitation during 1982-2005 and their changes over the twenty-first century. Temperature sensitivity of NPP in ESMs was generally consistent across northern high-latitude biomes but significantly more negative for tropical and subtropical biomes relative to satellite-derived estimates. The temperature sensitivity of NBP in both inversion-based and ESM estimates was generally consistent in March-May (MAM) and SeptemberNovember (SON) for tropical forests, semiarid ecosystems, and boreal forests. By contrast, for inversion-based NBP estimates, temperature sensitivity of NBP was nonsignificant for June-August (JJA) for all biomes except boreal forest; whereas, for ESM NBP estimates, the temperature sensitivity for JJA was significantly negative for all biomes except shrublands and subarctic ecosystems. Both satellite-derived NPP and inversion-based NBP are often decoupled from precipitation, whereas ESM NPP and NBP estimates are generally positively correlated with precipitation, suggesting that ESMs are oversensitive to precipitation. Over the twenty-first century, changes in temperature sensitivities of NPP, Rh, and NBP are consistent across all RCPs but stronger under more intensive scenarios. The temperature sensitivity of NBP was found to decrease in tropics and subtropics and increase in northern high latitudes in MAM due to an increased temperature sensitivity of NPP. Across all biomes, projected temperature sensitivity of NPP decreased in JJA and SON. Projected precipitation sensitivity of NBP did not change across biomes, except over grasslands in MAM.

Broad Consistency Between Satellite and Vegetation Model Estimates of Net Primary Productivity Across Global and Regional Scales

“…Despite the consistent drought response signs of the NPP S and NPP M across a large proportion of vegetated land areas, we found that the NPP M is generally more sensitive to drought than NPP S (Figures ), which is consistent with the result of a recent study in which drought was shown to cause an approximately 6 times larger reduction of CMIP5 model‐derived GPP compared with observation‐based GPP globally (Huang et al, ). Liu et al () found that CMIP5 NPP is oversensitive to seasonal precipitation variations relative to satellite NPP. Piao et al () also found that GPP from models is more sensitive to precipitation than data‐oriented GPP across biomes and globally.…”

“…Recently, long‐term satellite‐based NPP products have become available, such as satellite‐derived NPP (Smith et al, ), and they have enabled evaluations of the regional and global impacts of drought over the past three decades. For instance, Liu et al () used this satellite NPP product to evaluate the seasonal responses of terrestrial NPP to climate variation in the Coupled Model Intercomparison Project phase 5 (CMIP5) models.…”

Examining Forest Structure With Terrestrial Lidar: Suggestions and Novel Techniques Based on Comparisons Between Scanners and Forest Treatments

“…Forests are important for a variety of ecosystem services such as biodiversity, regulating climate, carbon storage, and hydrologic cycling but are being threatened on multiple fronts, including changes in land use, resource management, and climate change (Foley et al, 2005;Hansen et al, 2013;Trumbore et al, 2015). Temperate forest is one of the largest terrestrial carbon sinks on the Earth (Pan et al, 2011), but the sensitivity of the carbon cycle in these forests to changes in climate is uncertain (Liu et al, 2017). Changes in climate combined with altered forest structure and ecological function have led to increased frequency, size, and severity of wildfires, especially throughout western North America (Miller et al, 2008;Stephens et al, 2013; scale, spatially explicit descriptions of forest structure, condition, and change are necessary (Boisvenue et al, 2016;Dale et al, 1998;Dassot et al, 2011).…”

Variable tree rooting strategies improve tropical productivity and evapotranspiration in a dynamic global vegetation model

“…most roots are distributed downwards to a few meters of depths at maximum (Arora and Boer, 2003;Best et al, 2011;Guimberteau et al, 2017;Lawrence et al, 2011;Ostle et al, 2009;Schaphoff et al, 2018;Smith et al, 2014). By ignoring natural local adaptations of rooting depth, DGVMs and LSMs in the past had problems reproducing the extent of South-America's tropical evergreen forests, as well as its seasonal productivity and ET especially in regions with seasonal rainfall (Baudena et al, 2014;Liu et al, 2018Liu et al, , 2017Restrepo-Coupe et al, 2017).…”