| With the influence of global climate change and human activities, drought events occurred frequently and cause significant damage in ecosystem, natural environment and social economic development. Hydrologic drought, which takes the streamflow as main analysis factor, is related to meteorological drought and agricultural drought. Therefore, the assessment of hydrologic drought is fundamental to basin water cycle, water resources rational allocation and the protection against drought.The streamflow data covering 1956~2000 from four hydrologic stations in the upper reaches of the Fenhe River basin were used as case studies. First, selecting the the most suitable probability distributions for streamflow series with various time scale based on mathematical statistics methods. And the most suitable probability distributions were chose from normal distribution(NOR), exponential distribution(EXP), two-parameter log normal distribution(LN2), two-parameter gamma distribution(GAM2), pearson type ? distribution(P-?), log-pearson type ? distribution(LP-?), weibull distribution(W), generalized extreme value distribution(GEV), extreme value type I distribution(EV1), generalized pareto distribution(GPA), logistic distribution(LO) and generalized logistic distribution(GLO).Second, the hydrologic drought index was calculated by the most suitable probability distributions of streamflow series and the drought degree was divided. Last, the approaches of linear trend, mann-kendall, rescaled range analysis, Hilbert-Huang transform and markov were used to analyze the characteristic of hydrologic drought series and forecast the probability of each drought degree in the upper Fenhe River. The main content and results of this research were as follows:(1) The most suitable distribution selection for streamflow series with various time scales in the upper of Fenhe River.At month time scale, the GPA distribution was the best for streamflow series at Shangjingyou and Lancun hydrologic station. The GEV distribution was best for streamflowseries at Fenhe reservoir hydrologic station, the LP-III distribution was best for streamflow series at Zhaishang hydrologic station. At seasonal time scale, the W distribution was the best for streamflow series at Shangjingyou, Zhaishang and Lancun hydrologic station. The LP-III distribution was most suitable for streamflow series at Fenhe reservoir hydrologic station. At year time scale, the streamflow series at the four hydrologic stations all obey the GEV distribution.(2) The hydrologic drought index calculation and drought degree dividing in the upper of Fenhe River.The hydrologic drought index with various time scales was calculated by standardizing the most suitable cumulative distribution function. Three projects of drought degree dividing were taken into account in this study. The first one was the drought degree of standardized precipitation index(SPI). The second one was the drought degree divided by the level of high and low streamflow. The third one was the drought degree divided by the characteristics of standard normal distribution. Comparing with the real drought records of Fenhe River basin,the third project was the best one to assess hydrologic drought in the upper reaches of Fenhe River basin.(3) The analysis of each drought degree with various time scales in the upper reaches of Fenhe River.Severe drought was mostly occurred in January. Mild drought and moderate drought were mostly occurred in February. Because of snow melting, the non-drought and mild drought were mostly occurred in March to June. The time from July to October was flood season in Fenhe River basin, so non-drought and mild drought were most existed in these months. Mild drought and moderate drought were mostly occurred in November. Mild drought, moderate drought and severe drought were most appeared in December in so much as streamflow recharge source was insufficient. Mild drought mainly occurred in spring.Non-drought was mostly existed in summer. In autumn, the frequency of mild drought and non-drought were larger than other drought degree. In winter, mild drought, moderate drought and severe drought were all existed. Mild drought degree was mainly occurred at year time scale. At month time scale, consecutive non-drought, mild drought and extreme drought occurred for three months averagely in the upper reaches of the Fenhe River. For moderate drought and severe drought, the consecutive time duration was two to three months. At the seasonal scale, the consecutive time duration of each drought degree was almost two seasons.At year scale, the consecutive time duration of non-drought, moderate drought and severedrought was two years. The consecutive time duration of mild drought was about three years.And the extreme drought was not easy to occur continuously in the time scale of year.(4) Hydrologic drought character analysis in the upper reaches of Fenhe River.The mutation time of hydrological drought series with various time scales in the upper reaches of the Fenhe River was almost occurred in the year of 1970. The smaller the time scale was, the more the drought series period was. The greater the time scale was, the less the drought series period was. When the time scale was extended to year, there might be four drought series period range, which were two to three years, four years, five to seven years and twenty two years, in the upper Fenhe River. The hydrologic drought series in the upper Fenhe River decreased as time increased especially after the year of 1970. And the decreased trend of hydrological drought series had positive continuity.(5) Prediction of occurrence probability of each drought degree in the upper Fenhe River.When the initial state was non-drought or mild drought, the probability of non-drought or mild drought, which occurred again in one to six month in the upper Fenhe River, was greater than other drought degree. If the initial state was moderate drought or severe drought, the probability of them was greater than other drought degree in one to two month in the future and mild drought degree was mostly occurred in the next three to six month. When extreme drought was the initial state, it had biggest probability occurred again in the next one month.With the extension of prediction time, the probability that extreme drought transferred to other drought degree would increase. |
PDF Full Text Request