Spatial And Temporal Characteristics Of Droughts And Floods In The Hanjiang River Basin Since 1426

The Hanjiang River is one of the major tributaries of the Yangtze River.Located in the northern subtropical climate zone to the south of the Qinling Mountains and in the transition zone between China’s northern and southern climates,it is sensitive to climate and environmental change.As the water source area of the Middle Route of the South-to-North Water Diversion(MSNWD)project and a key river basin for flood control in the middle Yangtze River,the drought and flood(DF)control and water supply in this basin are of great importance in the national socio-economic development and ecological restoration.However,there is a lack of understanding of multi-decadal to century-scale drought/flood variability and extreme DF variability in the Hanjiang River basin(HRB);and there is still great uncertainty in the correlation between precipitation/droughts of North China(NC;water receiving area of the MSNWD project)and the Upper Hanjiang river(UH;water source area of the MSNWD project)at inter-decadal,multi-decadal and centennial scales.The establishment of historical DF grade series in the HRB and the study of the characteristic patterns of DF and extreme DF events in the HRB over the past centuries will not only improve scientific understanding of the decadal,multidecadal to century-scale variability of the frequency of precipitation and DF variability in this basin and its mechanisms,it will also contribute to the ability to predict DF variability on a more than decadal scale.Furthermore,it can provide basic scientific information for the construction,operation,and management of China’s major water conservancy project,the MSNWD project.In this study,we reconstructed the HRB’s DF series from 1426–2017 using a five-grade classification based on DF records from historical documents and precipitation data from the instrumental period,and identified extreme historical drought and flood events in this basin.Then the features of temporal variation,spatial distribution,and cyclical variation of DF and extreme DF in the HRB over 592 years were analyzed,respectively,and the factors influencing the variation of DF and extreme DF were also discussed.Furthermore,we analyzed the historical co-drought probabilities of the water source area(UH)and the receiving area(NC)of the MSNWD project on multiple time scales from 1470–2017 under the perspective of sustainable development of this project.The main studies and findings of this study are as follows:(1)By collecting and collating local chronicles and several published collections of Chinese historical documents on DF,we built the most comprehensive database on historical DF in the HRB(1426–1911).(2)The HRB has roughly experienced two relatively dry climate stages lasting around 100 years(from the early 15 th century to the early 16 th century and from the early 20 th century to the present)and one relatively wet climate stage lasting around four centuries(from the early 16 th to the early 20 th centuries)in the last 592 years.DF in the HRB became more frequent,and climate instability increased since the early 20 th century.The characteristics of the spatial distribution of DF showed that the UH,located in mountainous areas,has the greatest variability of DF.However,the high incidence of DF was concentrated in the middle and lower basin(ML),with the high frequency of floods centered around the downstream estuary and the high frequency of droughts centered around the downstream estuary and in the middle reaches.There is a phenomenon of interannual quasi-cycles of 2–8 for DF in the whole HRB as well as ML,and interdecadal to century-scale quasi-cycles of 10–30,50 and 80–100 years,and the upper HRB is dominated by 3–5 year quasi-cycles,as well as 10–30 year and 70–80year quasi-cycles.(3)From 1426–2017,a total of 45 extreme drought events and 52 extreme flood events occurred in the HRB,equivalent to an extreme drought event per 13 years and an extreme flood event per 11 years;and a total of 42 extreme drought events and 50 extreme flood events in the UH,corresponding to an extreme drought event per 14 years and an extreme flood event per 12 years.The relatively more frequent extreme drought events in the HRB occurred in the 15 th century to the early 16 th century,the 17 th century,and the 20 th century,and the highest frequency of extreme drought events occurred in the 20 th century;more frequent extreme flood events in the HRB occurred in the 16 th to 17 th centuries and the 19 th to 20 th centuries,and the highest frequency of extreme flood events was in the 19 th and 20 th centuries.The relatively more frequent extreme drought events in the UH occurred from the first half of the 17 th century to the end of the 17 th century,and from the second half of the 19 th century to the present,the frequency of extreme drought events was significantly higher in the 20 th century than in any other period;the concentration of extreme flood events in the UH was high during the 19 th and 20 th centuries.Overall,the frequency of both extreme DF events was low in the 18 th century,and the 20 th century saw a high frequency and increased risk of both extreme DF events in the HRB.(4)The multi-year intergenerational variation in the frequency of extreme DF events in the HRB is influenced to some extent by changes in the strength of the Asian summer monsoon,with the HRB being more prone to extreme drought events when the monsoon is weak and to extreme flood events when the monsoon is strong.On multi-year intergenerational scales,strong ENSO events,especially strong El Ni?o events,and large volcanic eruptions may impact annual and summer precipitation in the HRB,and in contrast,these two factors are more likely to play an influential role in extreme floods.Furthermore,AMO is significantly correlated with DF variability on interdecadal time scales in the HRB,the UH,and the ML;when AMO is in the positive phase,the HRB is relatively drought-prone to some extent overall.(5)Correlation of the DF grades between NC and the UH are significantly correlated on interannual,interdecadal,and multi-year intergenerational time scales,with the strongest correlations from the second half of the 16 th century to the middle of the 17 th century,while the longest duration of positive correlation is from the end of the 19 th century to the second half of the 20 th century.NC experienced the highest frequency of severe droughts from the second half of the 16 th century to the mid-17 th century and from the mid-19 th century to the present,while the UH in the same period also experienced a relatively high frequency of severe drought.During the driest 50-year period in NC from 1470–2017,there was a 52% probability that the UH was also drought-prone,with a 40% probability of a more severe drought event.During the three typical major drought events in historical periods in NC,i.e.,1637–1643,1877–1878and 1997–2002 megadroughts,the UH was also quite dry.In the study period’s seven driest years in NC,i.e.,1640,1641,1832,1920,1965,1968,and 1997,only 1832 was relatively wet in the UH,and 1968 was mildly dry,while the rest of the years were at or above Grade 4.Furthermore,EEMD and wavelet analysis indicate that NC and the UH have distinct cycles of DF variability of around 3–6 years,10 years,30,and 55-80 years,corresponding to the DF variabilities at interannual,interdecadal,and multi-decadal time scales.Overall,this study established a new historical data set of historical DF in the HRB,reconstructed historical DF series in the HRB with longer time scales and more detailed spatial resolution,identified the years of historical extreme DF,and found significant co-drought phenomena in NC and the UH at interannual,interdecadal and multi-decadal time scales.These efforts improved the understanding of the patterns of extreme DF variability at multiple time scales in the HRB.In the future,more historical documents,as well as other proxy data,need to be collected and used to improve the spatial coverage of basin data,explore the causal mechanisms of historical DF variability,and evaluate the social impacts caused by extreme historical DF variability in the HRB.