Temporal and spatial evolution patterns and prediction of drought in China in recent 500 years

: Drought is a prominent disaster in Chinese history. Analysing the spatial 26 and temporal evolution laws of drought could provide decision supports for drought 27 prevention and control. However, fewer studies were applied to investigate the 28 long-term evolution rules of drought events on different spatial and temporal scales. 29 In this paper, the spatio-temporal changes patterns of drought in China were analysed 30 with geostatistical methods based on the 1470–2000a drought datasets in China, and 31 then the possible future drought trend was predicted. Results showed that :(1) The 32 drought risk in the northern region was the highest during the past 500 years. And, the 33 drought intensity index showed an overall increasing trend with detail pattern of 34 weakening->strengthening-> weakening-> strengthening; (2) The drought condition 35 in the north was server than that in the south, but the drought trend in the south was 36 significantly aggravated. (3) The drought gravity centres were mainly distributed in 37 the north, but it showed a tendency to move southward. (4) From 1470 to 2000, the 38 study area showed a significant drought enhancements, which was predicted to show 39 an increasing trend of drought after 2000. 40


*Corresponding Author E-mail: 24
Email:guobingjl@163.com;zangwq@radi.ac.cn 25 Abstract: Drought is a prominent disaster in Chinese history. Analysing the spatial 26 and temporal evolution laws of drought could provide decision supports for drought 27 prevention and control. However, fewer studies were applied to investigate the 28 long-term evolution rules of drought events on different spatial and temporal scales. 29 In this paper, the spatio-temporal changes patterns of drought in China were analysed   The study area, located between 18.1°N-45.2°N, 95.9°E-128.3°E,mainly 109 includes the Liaohe River Basin, the Haihe River Basin, the Huaihe River Basin, the 110 Ye llow River Basin, the Yangtze River Basin (excluding the Shigu River Basin of Jinsha River), the Southeast River Basin, the Pearl River Basin and the Qinghai Lake 112 Basin ( Figure 1). The climate of the study area is largely affected by the monsoon, 113 with the precipitation ranging 200-1600mm.In addition, the temporal distribution of 114 precipitation differ greatly, which mostly occurs in summer. There is a decreasing 115 trend of precipitation from southeast to northwest in spatial distribution. The aridity in 116 northern part is larger than that of southern part.     (6) 168 When n ≥ 10, the statistic S approximately obeys normal distribution, and its does not consider the equivalent data points in the sequence.

170
The variance of the statistic S is The standardised test statistic Z is calculated as follows:  , and the logarithm can be obtained as follows: Where c is a constant, log n is the independent variable. The least square (1) H=0.5 indicates that the sequence is a random process, and the current trend 196 does not affect the future.

197
(2) 0<H<0.5 means that the time series presents anti-persistence, that is, the 198 future trend is opposite to the past. The closer the H value is to 0, the stronger the 199 anti-persistence will be.

200
(3) 0.5<H<1 indicates that the time series presents continuity, that is, the future 201 trend is the same as the past. The larger the H value is, the more obvious the trend is. Basin. In addition, the extent of high drought risk decreased during this period.

229
(3) During 1670-1769a, the drought risk had been further reduced. The mean 230 value and standard deviation of drought risk in this period were smaller than those 231 in the previous period, and the mean value was the smallest among the five periods.  Table 3 and Figure 3.

256
(1) During 1470-1519a, the drought risk was generally higher. The maximum 257 value, mean value and standard deviation in this period were the largest among the 10 258 periods, indicating larger spatial heterogeneity of drought risk inner the study area. As 259 shown in Figure 3(a), the drought risk was relatively higher in the northwest of the study area, mainly distributed in the Qinghai Lake Basin, the upper reaches of the 261 Yangtze River and the Yellow River Basin.

262
(2) During 1520-1569a, the drought risk decreased to some extent. The mean 263 value and standard deviation of drought risk were lower than those of the previous 264 period. As shown in Figure 3(b), regions with higher drought risk in the study area 265 were distributed in the Hetao Plain, the Liaohe River Basin and the Hongze Lake Haihe Plain had a slightly higher drought risk than the other areas.

292
(9) During 1870-1919a, the drought risk in the study showed an increased trend.

293
The standard deviation of drought risk was smaller, whereas the mean value was 294 larger. As shown in Figure 3(i), the zones with higher drought risk were mainly 295 concentrated in the west of the Loess Plateau and the east of the Pearl River Basin.

296
(10) During 1920-1969a, the drought risk had been further increased. In this 297 period, the mean value of drought risk was the largest, whereas the standard deviation 298 was the smallest among the 10 periods. This indicated that the spatial heterogeneity.

Variation characteristics of drought intensity
The drought intensity index can reflect the overall drought degree of the study 305 area. The larger the index value is, the more severe the drought will be. To better 306 analyse the characteristics of inter-annual variation of drought intensity, this study 307 calculated the drought intensity index of different study periods. The closer the 308 drought intensity index is to 3, the lower the drought intensity index is and vice versa.

309
As shown in Figure 4, the drought intensity showed an overall trend of fluctuating

Spatial distribution characteristics of average drought degree 328
The spatial distribution of average drought degree from 1470a to 2000a was 329 calculated. As shown in Figure 6, the drought in northern parts was the most serious

504
(2) The spatial distribution of drought was uneven during the past 500 years. The 505 drought intensity in the north was larger than that in the south, which showed a 506 significantly increasing trend.

507
(3) The gravity centres of drought were mostly distributed in Nanyang city, Overview of the study area and distribution of statistical sites. Note: The designations employed and the presentation of the material on this map do not imply the expression of any opinion whatsoever on the part of Research Square concerning the legal status of any country, territory, city or area or of its authorities, or concerning the delimitation of its frontiers or boundaries. This map has been provided by the authors. the legal status of any country, territory, city or area or of its authorities, or concerning the delimitation of its frontiers or boundaries. This map has been provided by the authors.    Spatial distribution of average annual drought degree in recent 500 years. Note: The designations employed and the presentation of the material on this map do not imply the expression of any opinion whatsoever on the part of Research Square concerning the legal status of any country, territory, city or area or of its authorities, or concerning the delimitation of its frontiers or boundaries. This map has been provided by the authors. Distribution of drought gravity centre from 1470 to 2000. Note: The designations employed and the presentation of the material on this map do not imply the expression of any opinion whatsoever on the part of Research Square concerning the legal status of any country, territory, city or area or of its authorities, or concerning the delimitation of its frontiers or boundaries. This map has been provided by the authors. The designations employed and the presentation of the material on this map do not imply the expression of any opinion whatsoever on the part of Research Square concerning the legal status of any country, territory, city or area or of its authorities, or concerning the delimitation of its frontiers or boundaries. This map has been provided by the authors.

Figure 9
Migration trajectory of the drought gravity centres in different time scales.