Climate Extreme Change And Its Relationship With Vegetation Growth And Productivity In Southwest China

Due to global warming,climate extremes are increasing significantly,which is greatly impacting vegetation dynamics.Southwest China（SWC）is identified as a key ecological area in China and has experienced frequent extreme climatic events.Using daily temperature and precipitation data from 119 meteorological stations and atmospheric circulation indices from1980 to 2018,self-calibrating Palmer Drought Severity Index from 2009 to 2010,and Moderate-Resolution Imaging Spectroradiometer data from 2000 to 2018,we adopted linear trend method,multiple stepwise regression method,and Pearson correlation coefficient to analyze the spatiotemporal changes of 21 extreme climate indices（ECIs）,normalized difference vegetation index（NDVI）,leaf area index（LAI）,and gross primary production（GPP）,as well as the responses of these metrics to ECIs,and calculate the resistance and recovery of the vegetation to extreme drought during 2009–2010.The main conclusions were as follows:（1）Except for consecutive wet days（CWD）,all wetness-related indices showed an increasing trend at different rates from 1980 to 2018.The temporal trends of precipitation-related ECIs（P-ECIs）showed that heavy precipitation days and extreme precipitation intensity in SWC had been increasing in recent decades.Spatially,the changes of P-ECIs varied with different regions.Due to extreme precipitation intensity increase,Guangxi was likely to suffer from flood disaster.However,in the Yunnan-Guizhou Plateau and the Hengduan Mountains,consecutive dry days（CDD）increased significantly（P<0.05）and CWD decreased significantly（P<0.05）,which indicated that drought may occur in these areas in the future.（2）The temporal trends of temperature-related ECIs（T-ECIs）indicated that SWC was becoming warmer.The warm-related indices and other indices increased,while the cold-related indices showed an opposite trend from 1980 to 2018.Among all T-ECIs,warm nights（TN90p）increased greatly at a rate of 0.32 days year^-1（P<0.001）,while cool nights（TN10p）decreased significantly(-0.32 days year^-1,P<0.001).The spatial distribution of T-ECIs showed that the warming in SWC mainly occurred in the high-altitude areas.（3）In SWC,the vegetation greenness and productivity were increasing as reflected by a significant increasing of NDVI,LAI,and GPP from 2000 to 2018 with an annual rate of 0.003,0.04 m² m^-2,and 10.58 g C m^-2,respectively（P<0.001）.The spatial distributions of the mean values of annual NDVI,LAI,and GPP were similar in most places,the high values were mainly at the edge of the Sichuan Basin,east of the Yunnan-Guizhou Plateau,and southwest Yunnan,the low values were observed in the Sichuan Basin and northwest Sichuan.The variation coefficients of NDVI,LAI,and GPP had higher values in the Hengduan Mountains,and the lower values in southwest Yunnan and Guangxi,indicating that the vegetation varied greatly in high-altitude regions.（4）Different ECIs had different,even opposing impact on the same vegetation metrics.Among the 9 P-ECIs,NDVI,LAI,and GPP were sensitive to CDD and heavy precipitation days（R10mm）.R10mm could promote NDVI,LAI,and GPP,while CDD showed negative impact on NDVI,LAI,and GPP.For T-ECIs,TN90p,max Tmin（TNx）,and diurnal temperature range（DTR）had the greatest impact on NDVI.Min Tmin,TNx,and growing season length had the greatest impact on LAI.GPP was sensitive to min Tmax,cool days,TN10p,and DTR.The sum of the relative contribution proportions of sensitive T-ECIs to NDVI,LAI,and GPP were 22.48%,12.98%,and 32.70%,respectively;this was higher than those of the sensitive P-ECIs（14.60%,12.75%,and 16.37%,respectively）.This indicated that T-ECIs had a greater impact on vegetation than P-ECIs in SWC.（5）Most ECIs were significantly correlated with NDVI,LAI,and GPP,with a time lag of1–3-month in SWC（P<0.05）.For P-ECIs,max 1-day precipitation amount and max 5-day precipitation amount had a significant correlation with NDVI and LAI with a lag of 3-month（P<0.01）,and had a 2-month lag for GPP（P<0.01）.Among T-ECIs,warm-related ECIs and other ECIs had a 1–3-month lag for NDVI,LAI,and GPP（P<0.05）.（6）During the 2009/2010 drought,the impacts of drought in different seasons on NDVI,LAI,and GPP were different.NDVI,LAI,and GPP were affected greatly by drought in spring,while less affected in autumn.Besides,among all the vegetation,deciduous broad-leaved forest was susceptible to drought,while evergreen broad-leaved forest had a higher resistance to drought.