Spatial-temporal Analysis Of Different Dry-Hot Scenarios In Yellow River Basin And Their Effects On Vegetation

Under the background of climate warming,the frequent occurrence of drought,extreme hot and compound dry-hot events has triggered a series of global ecological problems.In recent years,the international community has continued to pay more attention to drought,extreme hot and compound dry-hot events.In order to effectively identify the occurrence and development process of drought,extreme hot and compound dry-hot events,domestic and foreign scholars have constructed a large number of drought and extreme hot indices based on information such as vegetation index,temperature,precipitation,and evapotranspiration.However,the spatial-temporal resolution of existing drought and extreme hot datasets is still insufficient.First of all,the drought and extreme hot datasets constructed based on meteorological stations are spatially discontinuous.In addition,the drought and high temperature datasets constructed using reanalysis data have low spatial resolution.This leads to many uncertainties in monitoring and forecasting research on drought and extreme hot events.Based on these problems,this paper selects the Yellow River Basin(YRB)as the research area,and builds a Random Forest Regression(RFR)model based on meteorological data,remote sensing data and reanalysis data to produce drought and high temperature index with high spatial-temporal resolution.A high-spatial-resolution drought and high temperature index dataset of the YRB from 2001 to 2020 was produced,which effectively revealed the spatio-temporal variation of drought and extreme hot events in the YRB in the past 20 years.Secondly,in order to accurately evaluate the impact of different dry-hot scenarios on vegetation growth,this paper uses the Vine Copula Conditional Probability(VCCP)model to clarify the probability of vegetation loss under different drought,high temperature,and combined dry-hot scenarios.Finally,this paper clarifies the spatio-temporal variation of climate factors such as temperature,precipitation and potential evapotranspiration in the YRB,and discovers the vulnerability of different vegetation types under different dry and hot conditions.The main conclusions are as follows:(1)A new and robust high spatial-temporal resolution Standardized Precipitation Evapotranspiration Index(SPEI)and Standardized Temperature Index(STI)model was constructed.Based on the SPEI and STI models,the high spatial resolution(1 km)SPEI and STI datasets of the YRB from 2001 to 2020 were produced on a 3-month scale per month.The R2,Mean Error(ME),Mean Absolute Error(MAE) and Root Mean Square Error(RMSE)of the SPEI and STI datasets on the training samples were 0.944,-0.02,0.172,0.259 and 0.981,-0.006,0.0991,0.207,respectively;On the testing samples,the R2,ME,MAE and RMSE were 0.653,-0.01,0.5,0.66 and 0.762,-0.04,0.575,0.77,respectively.(2)The spatial-temporal variation of SPEI,STI,and NDVI in the YRB from 2001 to 2020 was clarified.In the past 20 years,the SPEI,STI and NDVI in the YRB have all shown an upward trend,that is,the drought situation has gradually weakened,the temperature has gradually increased,and the vegetation growth has gradually improved.The areas where SPEI,STI,and NDVI showed significant increases were 37.11×104km2,18.83×104km2,and 68.32×104km2,accounting for 46.66%,23.67%,and 85.89% of the total area of the YRB,respectively.(3)The interaction mechanism among SPEI,STI and NDVI in the YRB is clarified.The area affected significantly(P<0.05)by drought(SPEI)on vegetation(NDVI)in the YRB(24.59×104 km2)was much larger than the area affected by STI(3×104 km2).The results of the correlation between SPEI and STI index show that the SPEI and STI are positively correlated in the west of the YRB(high altitude area,24.78×104 km2),that is,the drought condition will decrease as the temperature rise(increased meltwater from glacial and snowpack).However,in the central and eastern regions(54.81×104 km2),the SPEI is negatively correlated with STI,that is,the drouth will increase as the temperature rise.(4)The probability of drought,extreme hot,and combined dry-hot events affecting vegetation was quantified.When the drought intensifies,the area where the probability value of the NDVI index drops below the 0.5th percentile exceeds 60% increases significantly;when the temperature rises,the area where the probability value of the NDVI index drops below the 0.5th percentile exceeds 60% has a slightly increase trend.From the perspective of vegetation loss probability,drought intensification has a greater stress intensity on vegetation than temperature increase.Under combined dry-hot conditions,vegetation growth is mainly subject to water stress,and the aggravation of drought will directly destroy the normal growth of vegetation.(5)The vulnerability of different vegetation types under different dry and hot conditions is revealed.The dry and hot events have the least impact on the vegetation in the wet and cold area of the YRB,but the greatest impact on the vegetation in the dry-cold area.Under non-drought conditions,grassland and cropland growth was less inhibited than mixed forest,deciduous forest,and evergreen forest growth,while under drought conditions,grassland and cropland growth was more inhibited than mixed forest,deciduous forest,and evergreen forest.Grasslands in dry-hot,humid-hot and dry-cold areas,and farmland in dry-cold and humid-cold areas are the most vulnerable vegetation types in the YRB.