The Variations Of Terrestrial Gross Primary Production In China And Its Responses To Drought

Drought can substantially alter the composition,structure and function of terrestrial ecosystems,thus significantly affecting terrestrial carbon cycle.Gross primary production(GPP),reflecting the productivity level of terrestrial ecosystems,is the largest component of the global terrestrial carbon fluxes.Here we investigated the spatio-temporal patterns of GPP and drought in China during 1979–2015,and quantitatively analyzed the responses of GPP to drought at different time scales.The GPP data used in this study included four terrestrial ecosystem models(TEM)(DLM,CABLE,LPJ and ORCHIDEE)simulated GPP,eddy-covariance measurements based upscaling GPP(FLUXCOM),and satellite remote sensing GPP(MODIS).The drought data included drought indexes: Standardized Precipitation Evapotranspiration Index(SPEI)and self-calibrating Palmer Drought Severity Index(sc PDSI).The major conclusions were as follows.Firstly,the GPP showed a downward gradient from southeast to northwest in China during 1979–2015.The abnormal changes in GPP at different time scales were obvious and the large interannual variations mainly occurred in the northeast forest area;Most parts of China had shown a clear upward GPP trend,accounting for about half of China's total area,while the downward trend only accounted for about 5%.Secondly,the large-scale drought in China mainly occurred in the mid-1980 s,the mid-1990 s,the early 21 st century,and the last 5?6 years.At the interannual period,the average value of the arid regions showed a slightly decrease.The spatial distribution of SPEI and sc PDSI at 8 different time scales was consistent;the overall trend was that the western regions were becoming moist,and the northeast regions were drying.Thirdly,the time-variation curves of GPP and drought index were consistent at different drought time scales.The SPEI drought index at shorter time scale was higher than those at long-term time scales.The effect of summer drought on vegetation productivity was more significant than that in spring and autumn.With the increasing of SPEI drought index time scales,the correlation is higher.In addition,the GPP estimates positively correlated with SPEI in most regions of northern China,but negatively in many areas of southern China.Different GPP datasets had a good spatial consistency of sensitivity to SPEI.The higher sensitivity was mainly distributed in semi-arid areas such as the Junggar Basin,the Qinghai-Tibet Plateau,the Qaidam Basin,and the plains of the middle and lower reaches of the Yangtze River.At last,the GPP response to drought is closely related to the time scale of the selected drought index.The GPP's responses to drought response and sensitivity varied with different time scale.Overall,to obtain a more accurate estimation of GPP,there is an urgent need to improve the current GPP models and establish more flux observation.