| Under the background of continuous population growth,rapid economic development and abnormal global climate change,the extremity and uncertainty of climate change rises,extreme climate events are gradually increasing,and in many basins,the average climate,hydrological situation,extreme precipitation and runoff characteristic have changed significantly.In recent years,the frequency and influence range of drought in Luanhe River basin are on the rise,and some rivers are even interrupted.Therefore,the study of drought identification,spatio-temporal variation,driving mechanism and drought simulation and evaluation of future scenarios in the sphere of Luanhe River basin,is of great practical significance for the comprehensive response to future drought,scientific and effective drought control,as well as sustainable economic and social developmentTaking the Luanhe river basin as an example,based on the distributed hydrological model,this paper carried out the studies of the identification of historical drought characteristics,the evolution law,driving mechanism in the changing environment and the response of meteorological,agricultural and hydrological drought in the basin under the combined action of climate change and land use change.The above research is expected to provide references for scientific management and rational planning of water resources in the basin and drought response.The main contents and conclusion are as follows:(1)By optimizing the probability distribution of monthly runoff of 7 hydrological stations in Luanhe River basin,the o ptimal standardized runoff index of each hydrological station was calculated.The trend characteristics of hydrological drought in different time scales were discussed.The spatial evolution characteristics of interannual and seasonal variations of drought characteristics and drought frequency were studied.It is found that the General Extreme Value and Log-Logistic distributions performed prominently in fitting the monthly streamflow of Luanhe River basin,while the common Gamma distribution did not fit well.The hydrological drought had an obvious aggravating trend over the past 51 year.In the context of climate change and global warming,the trend of hydrological drought in the study area is likely to continue,that is,the drought situation may worsen in the future.The drought in the Qianan+Lulong+Luanxian(QA+LL+LX)region and the junction region of Longhua(LH)and Luanping(LP)are was most severe.Compared with the mild and more serious drought,the severe and more serious drought has the characteristics of short duration and heavy damage.The whole basin suffered the mild and more serious drought,while the severe and more serious drought only cover some areas.(2)Based on the hydrological simulation of SWAT model and continuous 3-dimensional space-time drought identification method,the hydrological drought events in the Luanhe River basin from 1961 to 2011 were identified.The spatial and temporal variation of the three most severe drought events was reconstructed by using 3-dimensional visualization technology.The joint probability characteristics of drought duration,affected area and drought intensity in Luanhe River basin were analyzed by multivariate frequency analysis method based on Copula function,and the recurrence period of typical extreme drought events was calculated.Results show that the result of drought recognition by the method of spatio-temporal continuous three-dimensional drought recognition is consistent with the historical drought record,which show the method of drought recognition used in this paper is reasonable and reliable.The west and middle and lower reaches of Luanhe River are the drought center of the basin,and large-scale drought events are mostly concentrated in this region.The optimal marginal distributions for S,D and A are GEV,GEV and Lognormal distribution,the optimal bivariate copula function for S-D,S-A and D-A are Joe Copula,Gumbel Copula and Joe Copula,and the optimal trivariate copula function for S-D-A is Gumbel copula.Compared with the symmetric Copula functions,the asymmetric Copula functions are more suitable for drought multivariate frequency analysis.The recurrence period of most serious drought event(June 2000 to June 2003)in the Luanhe River basin from 1961 to 2011 is 80 years.In addition,there was a drought once every 75 years and three droughts once every 40 to 50 years.In frequency analysis,regardless of any one of drought duration,affected area and drought intensity,the recurrence period of drought will be seriously underestimated,which is very unfavorable to drought risk management and the planning and design of drought-resistant water conservancy projects.(3)By selecting appropriate drought index(SPEI,SSI and SRI)and using SWAT model as the relationship between the index,the evolution characteristics between meteorological,agricultural and hydrological drought in the basin were discussed.The relationship between meteorological drought,agricultural drought,hydrological drought and large-scale climatic circulation factors was studied by using the methods of continuous wavelet and cross wavelet transform.The response of agricultural and hydrological droughts to the lack of precipitation showed a certain time lag,which became obvious with the increase of the drought scale.The evolution time between meteorological drought and agricultural drought,and agricultural drought and hydrological drought showed remarkably seasonal characteristics in Luanhe river basin.The meteorological drought has strong positive correlations with SOI,and negative correlations with Ni?o 1+2,Ni?o 3,Ni?o 4,Mixed Ni?o 1+2+3+4,Ni?o 3.4 and MEI events.And the overall maximum correlations with Ni?o 1+2,Ni?o 3,Ni?o 4,Mixed Ni?o 1+2+3+4,Ni?o 3.4,SOI and MEI occur at scale-9 and lag-0.The overall strongest correlations with AOM are obtained at the lag-6 and scale-24.The agricultural drought has a stronger positive correlation with Ni?o 4,Ni?o 3.4 and MEI,but a negative correlation with SOI.The correlations with Ni?o 4,Ni?o 3.4,SOI and MEI reach the overall maximum values at scale-17 with lag-7,scale-16 with lag-10,scale-16 with lag-9,and scale-18 with lag-9,respectively.The hydrological drought has strong negative correlations with AOM and Ni?o 4,and the correlations between Ni?o 4,AOM and SRI reaches an overall maximum at scale-10 with lag-17 and scale-4 with lag-5,respectively.The significant influence of the Ni?o 1+2+3+4,Ni?o 3.4,Southern Oscillation Index(SOI),Multivariate ENSO Index(MEI)and Atlantic Multidecadal Oscillation(AOM)on meteorological drought was concentrated in the 16-88-month periods,as well as the decadal scale of 99-164-month periods,the significant influence of Ni?o 4,Ni?o 3.4,MEI,and SOI on agricultural drought was concentrated in the 16-99-month periods,as well as the decadal scale of 99-187-month periods,and the significant influence of Ni?o 4 and AOM on hydrological drought was concentrated in the 16-64-month periods,as well as the decadal scale of 104-177-month periods.These results presented quantitative propagating linkages among three types of droughts and their related driving factors,which is significant for regional drought monitoring and strategic decision-making.(4)In order to reduce the uncertainty in climate change research,this paper adopts the multi-model climate change prediction method based on Bayesian weighted average(REV-BMA)to obtain ensemble GCMs data.Compared with the measured meteorological elements,the applicability of each single model and ensemble model in Luanhe River basin was evaluated.The precipitation and temperature series in the future(2011-2100)were obtained by SDSM downscaling and bias correction.The multi-model ensemble mean has the best hindcast.The simulation effect of SDSM model on precipitation is better than in the calibration period than the verification period,and the simulation effect on summer with the relatively abundant water is better than that in spring,autumn and winter.Especially,the simulation effect on winter is the worst due to the less precipitation in winter.The simulation effect of maximum and minimum temperature is obviously better than that of precipitation.The mean deviation of simulated precipitation was significantly reduced by the bias correction,and the distribution of the corrected precipitation data was basically consistent with that of the observed data.The simulation precision of SDSM model for temperature is high,and the mean and probability distribution of the maximum and minimum temperature is not improved significantly by the bias correction.However,this does not mean that the simulation accuracy of SDSM model for maximum and minimum temperature is low.Therefore,the SDSM model with bias correction can be used to construct the next climate scenario.(5)Land use in the future(2020s,2050 s and 2080s)was predicted by the CA-Markov model.The meteorological elements obtained from downscaling model and the land use predicted by CA-Markov model in the future period were taken as the input data of the SWAT model to simulate the soil water content and runoff changes in the future period in the Luanhe River basin.On this basis,combined with the run theory and empirical orthogonal(EOF)methods,the temporal and spatial variation characteristics of drought in Luanhe River basin in the future were revealed.The land use type predicted by the CA-Markov model has good reliability and applicability.In the future,cultivated land(dry land and paddy field)will continue to decrease,while the area of construction land will continue to expand on the existing basis,almost entirely relying on occupied farmland.Under the RCP2.6 scenario,the northwest region of WC,FN and LH,the west region of NM in the middle and upper reaches of the Luanhe River,and the north region of LY and JL+QL,and the XX+ZH region in the lower reaches,are prone to moderate and more serious meteorological drought.Under the RCP4.5 scenario,the region of WC,FN and NM,the south region of LY,KC and PQ,the southwest CD region,the east XL region,the northeastern JL+QL region and XX+ZH region are prone to moderate and more serious meteorological drought.Under the RCP8.5 scenario,the northeast boundary zone in the middle and upper reaches of Luanhe River(the northeast NM,east WC and the east LH region)is prone to moderate and more serious meteorological drought.Under three emission scenarios,the upper reaches of Luanhe River and the southwest region of the middle reach are prone to moderate and more serious agricultural drought in the future.Under the RCP2.6 scenario,the most region of southeast WC and FN,the most region of northwest LH and northwest LP in the middle and upper reaches of the Luanhe river,and the LY,north JL+QL and east PQ region in the lower reaches are prone to the hydrological drought.Under the RCP4.5 scenario,the FN,southwest NM,southeast WC,northwest LH and LP region in the middle and upper reaches of Luanhe River are prone to the hydrological drought.Under the RCP8.5 scenario,the southwest and southeast region of WC,southeast FN,the most LH,northwest LP,east CD,and the junction region of XL,KC and XX+ZH are prone to the hydrological drought.Compared with the RCP2.6 and RCP8.5 scenarios,the recurrence period of meteorological,agricultural and hydrological droughts in the RCP8.5 scenario is shorter,that is,drought is more likely to occur. |
PDF Full Text Request