论文部分内容阅读
Abstract Drought is one of the major meteorological disasters affecting the climate in China. In this paper， the interannual variation and seasonal distribution changes of drought at different time scales were analyzed with the standardized precipitation index （SPI） as the drought evaluation criterion to the precipitation data of the 5 meteorological stations of Jinan， Tai’an， Yiyuan， Shen County and Yanzhou from 1960 to 2013. The results showed that： ① the frequency of drought was low in spring and summer in inland areas of Shandong Province， while autumn had frequent occurrences of moderate drought， and winter had frequent occurrences of heavy and severe droughts. ② In the 1960s， 1980s and early 21st century， the number of droughts increased significantly， and the SPI values showed a significant decrease. ③ The 3month time scale range was wide， during which the frequency of occurrence was high. The trend of SPI changes at the 12month time scale was affected by the accumulation of antecedent precipitation， and the change was slow. The research results can provide a scientific reference for arid climate analysis and water resources management in agriculture and production in inland areas of Shandong Province.
Key words Standardized precipitation index; Inland areas of Shandong Province; Drought; Characteristics
Drought is one of the most important natural meteorological disasters in China. It occurs frequently， has a long duration， and has a wide range of impacts. The occurrence of drought is likely to cause great agricultural production reductions， serious desertification， and lack of water resources， posing great impacts on normal agricultural production and people’s lives. Due to the influences of monsoon climate and landforms， Shandong Province is characterized by uneven distribution of water and heat and unbalanced precipitation， resulting in frequent occurrence of drought disasters. Therefore， the study and analysis on the drought characteristics of Shandong Province， especially the inland areas of Shandong Province， is of great significance to understand the occurrence and development trend of drought， timely adjust agricultural production activities and reduce the losses caused by drought.
As for the characteristics of drought， some scholars have proposed the Palmer Drought Index （PDSI）[1]， the Surface Water Supply Index （SWSI）[2] and the Standardized Precipitation Index （SPI）[3]. Among them， SPI is widely used because of its simple calculation， flexible time scale and better reflection of drought characteristics. At present， SPI has been more and more used to analyze the temporal and spatial variation characteristics and seasonal characteristic changes of droughts at home and abroad[4-19]， so as to better study the characteristics of drought evolution. Piccarreta[20] used SPI to analyze the drought in southern Italy from 1923 to 2000. Guttmann[21] compared SPI and PDI based on the precipitation data of the United States， and found that PDI spectral characteristics changed with the locations， presenting high uncertainty， while SPI did not， and SPI was much easier to use than PDI. SPI is one of the main bases of drought detection system in the United States[22]. In China， Zhang et al.[23] used SPI to analyze and study the drought and flood conditions and their occurrence frequency in the drought wetness episodes in the Pearl River Basin. Zhang et al.[23] analyzed the characteristics of drought evolution in Gansu Province based on SPI. Yuan et al.[4] analyzed SPI and Z index in China， and concluded that SPI could effectively analyze the drought situation in China. It could clearly reflect the changes of drought characteristics and had a good predictive effect on drought disasters， making it better than Z index. In this study， based on the monthly precipitation data （from the China Meteorological Data Service Center） of 5 meteorological stations in central and western Shandong Province from 1960 to 2013， the precipitation characteristics， SPI change characteristics and drought characteristics in the inland areas of Shandong Province from 1960 to 2013 were analyzed with SPI as the meteorological drought analysis index in the time scale from the aspect of meteorological drought， so as to provide a theoretical basis for drought monitoring， early warning and prevention of drought disasters.
Materials and Methods
Study area overview
The inland areas of Shandong Province have a warm temperate monsoon climate with concentrated precipitation， rain and heat falling in the same season， short spring and autumn， and long winter and summer. The annual average temperature is 11-14 ℃. The inland areas of Shandong Province have abundant light resources with the annual average sunshine time of 2 290-2 890 h， and the heat conditions can meet the needs of biannual cropping. The annual average precipitation is generally around 65-70 mm. The seasonal distribution of precipitation is very uneven. About 60%-70% of the annual precipitation falls in July-September. Moreover， it is easy to have flooding disasters in summer， and drought disasters in winter， spring and late autumn， which has a great impact on agricultural production.
Data and methods
Data acquisition
All precipitation data used in this paper was provided by the China Meteorological Data Service Center， including the month by month precipitation data of the 5 major meteorological stations in Jinan， Shen County， Yanzhou， Yiyuan and Taishan from 1960 to 2013. Statistically speaking， such a long time series was sufficient enough to obtain a more credible trend result.
SPI method
SPI is an indicator of the probability of occurrence of precipitation in a certain period of time， and it is suitable for drought detection and assessment relative to the local climatic conditions at the monthly or multimonthly time scales. The calculation of SPI is based on the assumption that the precipitation is a kind of skewed distribution. The Γ distribution probability is used to describe the change of precipitation， and the precipitation of the skewed probability distribution is normalized. Finally， the drought level is classified by the cumulative frequency distribution of standardized precipitation. The SPI calculation formula is as follows： SPI=St-（c2t+c1t）+c0d3t+d2t+d1t+1.0（1）
Where， t=ln1G（x） 2， and G（x） is the probability of precipitation distribution associated with the Γ function; x is the annual or seasonal precipitation sample; S is the positive and negative coefficients of probability density.
If G（x）>0.5， then S=1; if G（x）≤0.5， then S=-12. G（x） is calculated by the probability distribution integral formula of the Γ distribution function：
G（x）=1β γΓ（γ）0∫ x0xγ·1e-x/βdx， x>0（2）
Where， γ， β are the shape and scale parameters of the Γ distribution function; c0， c1， c2 and d1， d2， d3 are the calculation formulas for the simplified approximate solution formulas of the Γ distribution function converted to the cumulative frequency， and c0=2.515 517， c1=0.802 853， c2=0.010 328， d1=1.432 788， d2=0.189 269， d3=0.001 308.
According to the SPI values obtained by the above formulas， the SPI values for different drought levels were determined according to the frequency of occurrence of drought （Table 1）[24].
Linear regression method
On the basis of a large amount ofprecipitation data， the SPI value was calculated， and the mathematical statistics method was used to establish the regression function expression between the dependent variables and the independent variables， which could show the variation trends of SPI in the inland areas of Shandong Province.
Results and Analysis
Basic characteristics of precipitation in inland areas of Shandong Province from 1960 to 2013
The comparison on the mean， maximum， and minimum values of precipitation （Table 2） showed that the precipitation in Taishan area was the largest， while that in the western part of Shandong Province was relatively small. This was mainly due to the foehn effect. The Taishan side was on the windward slope， and the humid airflow was forced to rise when encountering the highlands like mountains. Thus， the temperature dropped， forming rainfall. On the other hand， the precipitation in western Shandong Province was little because it lied in the Huangpan Plain， where the shallow surface water evaporation effect was poor， and the groundwater storage was low， making it impossible to achieve precipitation cycle. At the same time， due to the influence of sea and land factors， it was impossible to carry out the precipitation cycle between land and sea. Therefore， the uneven distribution of precipitation was prone to result in uneven harvest situations in the study areas. In order to analyze the seasonal variation of drought， the variances of winter， spring， summer and autumn were calculated separately[29]. The results show that the variances of winter， spring， summer and autumn were 0.113， 0.402 676， 0.312 823 and 0.319 217， respectively in the inland areas of Shandong Province from 1960 to 2013. Therefore， the variation in spring was the biggest. The variation curve of SPI in each season was shown in Fig. 3.
The drought situation was obvious in spring in the inland areas of Shandong Province， and the SPI values showed no significant fluctuation. Besides， the spring drought occurred in the years of 1969， 1974， 1976， 1981， 1992 and 2012， in which the spring drought in 1968 was the most serious with the smallest SPI value， while SPI of the other years tended to be normal. The results showed that from the longterm trend， the spring droughts slowly decreased， and drought damage gradually reduced in the inland areas of Shandong Province. The spring precipitation tended to be stable， and the change trend was not obvious.
There was almost no drought in summer in the inland areas of Shandong Province， and the SPI values also tended to be stable. It was easy to have rain floods in summer， and the occurrence times of extreme weather somewhat increased in some periods. The results indicated that from the longterm run， there was no obvious drought in summer in the inland areas of Shandong Province.
On the other hand， the droughts in autumn were serious in the inland areas of Shandong Province. Severe droughts occurred in 1980， 1996 and 2009， especially after entering the 21st century， when the occurrence of drought became frequent. In addition， the variation of drought levels tended to be stable， and it was common to have light or moderate droughts， indicating that the autumn had high occurrence frequency of droughts in the inland areas of Shandong Province. The results showed that in the longterm trend， the autumn drought occurred frequently in the inland areas of Shandong Province， and the precipitation distribution in autumn was more uneven with more significant changes. Gradually， the autumn became the raindeficient season.
The drought conditions were extremely serious in winter the inland areas of Shandong Province， and droughts occurred throughout the whole winter. The droughts were very serious in the 1960s， 1980s， and early 21st century. The variation of drought levels was greater during the period 1974-1998 than in other time periods， and three severe a total of 3 heavy drought occurrences. It indicated that the frequency of drought occurrence was greater in winter than in any of the other 3 seasons in the inland areas of Shandong Province. Moreover， the longterm trend showed that winter drought was serious in the inland areas of Shandong Province， which was characterized for the extreme water shortage and significant changes. Conclusion
In this paper， SPI values of different time scales are calculated based on the month by month precipitation data from 5 inland areas in Shandong Province from 1960 to 2013 by using SPI as the drought assessment factor， and analysis has been made to the overall characteristics of precipitation， SPI interannual variation and seasonal distribution changes in these areas in the 54 years. Thus， the following conclusions are drawn.
（1） The drought characteristics change significantly in the inland areas of Shandong Province from 1960 to 2013. There are few occurrences of drought in spring and summer， while autumn and winter are the highincidence seasons of droughts. Moreover， it is easy to have moderate droughts in autumn， and severe droughts in winter.
（2） There is a general trend of drought in the inland areas of Shandong Province， and with the increase of time scale， SPI becomes much less random but more continuous. The occurrences of droughts were frequent in the 1980s and 1990s， and drought has also become the major natural disaster in the 21st century.
（3） In terms of interannual variations in drought， spring drought was the most severe in the 1960s， autumn drought was most severe in the 1980s， and winter drought was the most severe in the 1990s.
（4） As for the seasonal distribution of droughts， the area of spring drought gradually decreases， and there is almost no drought in summer. The drought area in autumn increases and gradually becomes more serious， and the drought in winter shows no significant variation. In general， the drought in autumn is the most serious throughout the whole province.
The analysis on the interannual and seasonal distribution characteristics of drought in the inland areas of Shandong Province shows that local agricultural production should pay full attention to the drought in spring and autumn， especially the occurrence of autumn drought. It should strengthen and improve the farmland irrigation system in the inland areas of Shandong Province， so as to improve the overall levels of drought prediction， detection and control in the areas. In this paper， the interannual and seasonal distribution changes are analyzed only from the SPI values at different time scales in the inland areas of Shandong Province from 1960 to 2013， but no consideration has been given to the continuity and variability of the drought across different seasons. In addition， as a complex meteorology， drought is a process of comprehensive action of the atmosphere， hydrosphere and biosphere. In this paper， the characteristics of drought are analyzed and studied only from the effects of precipitation， without considering the joint influences of the hydrological processes of the surface， atmospheric circulation and human activities， which should be the focus in further studies. References
[1] PALMER WC. Meteorological drought research paper[J]. 1965 （3）： 62-67.
[2] SHAFER BA， DEZMAN LE. Development of a Surface Water Supply Index （SWSI） to assess the severity of drought condition s in snow pack runoff areas Proceedings of the Western Snow Conference[N]. Colorado： Colorado State University， Fort Collins， 1982： 164-175.
[3] MCKEE TB， DOESKEN NJ， KLEIST J. The relation of drought frequency and duration to times cales1 Proceedings of the Eighth Conference on Applied Climatology[N]. Boston： Amer Meteorological Society， 1993： 179-1841.
[4] YUAN WP， ZHOU GS. Comparison between standardized precipitation index and Zindex in China[J]. Acta Phytocologica Sinica， 2004， 28（4）： 523-529.
[5] XUE DQ， WANG JG.. Characteristic analysis of change into drought in Shandong Province[J]. Journal of Natural Disasters， 2007， 16（3）： 60-65.
[6] MA JY， XU YL， PAN J， et al. Drought tendency based on standardized precipitation index （SPI） and relative moisture over northeast China from May to September during 1961-2009[J]. Journal of Meteorology and Environment， 2012， 28（3）： 90-95.
[7] SIMS AP， NIYOGI DS， RAMAN S. Adopting drought indices for estimating soil moisture： a North Carolina case study[J]. Geophysical Research Letters， 2002， 29（8）： 1183-1186.
[8] HUANG WH， YANG XG， LI MS， et al. Evolution characteristics of seasonal drought in the south of China during the past 58 years based on standardized precipitation index[J]. Transactions of the Chinese Society of Agricultural Engineering， 2010， 26（7）： 50-59.
[9] CHE SJ， LI CQ， SHEN SH. Analysis of droughtflood spatialtemporal characteristics based on standard precipitation index （SPI） in Hebei Province during 1965-2005[J]. Chinese Journal of Agrometeorology， 2010， 31（1）： 137-143.
[10] LI SJ， YUAN J， XIAO QH， et al. Characteristics of droughts and humidification based on standard precipitation index in Weifang City during the past 50 years[J]. Chinese Agricultural Science Bulletin， 2011， 7（26）： 293-298.
[11] LI WG， CHEN HL， ZHU NH， et al. Analysis of drought monitoring on Hainan Island from standardized precipitation index[J]. Chinese Journal of EcoAgriculture， 2009， 17（1）： 178-182.
[12] VICENTESERRANO SM. Spatial and temporal analysis of droughts in the Iberian Peninsula （1910-2000）[J]. Hydrological Sciences Journal， 2006， 51（1）： 83-97.
[13] GONG HQ， TAI QG. Analysis of rainfall characteristics in Linyi City[J]. Journal of Anhui Agricultural Sciences， 2008， 36（ 28）： 12356-12357. [14] YANG YQ. Statistical methods in agricultural meteorology[M]. Beijing： Meteorological Press， 1983， 153-160.
[15] SEIER RA， HAYES M， BRESSAN L. Using the standardized precipitation index for flood risk monitoring[J]. Int J of Climatology， 2002（22）： 1365-1376.
[16] PATEL NR， CHOPRA P， DADHWAL VK. Analyzing spatial patterns of meteorological drought using standardized precipitation index[J]. Meteorological Applications， 2007（14）： 329-336.
[17] EDWARDS DC， MCKEE TB. Characteristics of 20th Century Drought in the United States at Multiple Time Scales[D]. USA： Department of Atmospheric Science， Colorado State University， 1997： 18-21.
[18] ZHANG Q， ZOU XK， XIAO FJ. GB/T 20481-2006 Classification of meteorological drought[S]. Beijing： China Standard Press， 2006： 16-17.
[19] VICENTESERRANO SM， CUADRATPRATS JM. Trends in drought intensity and variability in the middle Ebro valley during the second half of the twentieth century[J]. Theoretical and Aplied Clinatology， 2007（88）： 247-258.
[20] PICCARRETA M， CAPOLOCONG D， BENZI F. Trend analysis of precipitation and drought in Basilcate from 1923 to 2000 within a southern Italy context[D]. Int. Climate， 2004， 24： 907-922.
[21] GUTTMAN NB. Comparing the Palmer drought index and the standardized precipitation index[J]. Journal for the American Water Resources Association， 1998， 34（1）： 113-121.
[22] SWOBODA D， HAYES M. The Drought Monitor Amer Meteor Soe[J]. 2002， 83： 1811-1190.
[23] ZHONG Q， XU CY， ZHANG ZY. Observed changes of drought wetness episodes in the Pearl River Basin[D]. China， 2009， 98： 89-99.
[24] General Administration of Quality Supervision， Inspection and Quarantine of the People's Republic of China， China National Standardization Administration. GB/T20481-2006 Classification of meteorological drought[S]. Beijing： China Standard Press， 2006.
[25] WEI CY， WANG JJ， ZHANG F， et al. Using expansion of empirical orthogonal function and geostatistics to analyze temporalspatial variability of precipitation[J]. Journal of Irrigation and Drainage， 2010， 29（4）： 105-109.
[26] EDWARD DC， MCKEE TB. Characteristics of 20th Century Drought in the United States at Multiple Times Scales[J]. USA： Department of Atmospheric Science， Colorado State University， 1997 ：18-21.
[27] HUANG WH， YANG XG， LI MS， et al. Evolution characteristics of seasonal drought in the south of China during the past 58 years based on standardized precipitation index[J]. Transactions of the Chinese Society of Agricultural Engineering， 2006， 26（7）： 50-59.
[28] CHEN J， LU DH. Division of monsoon climatic regions in China[J]. Journal of Beijing Forestry University， 1981： 34-41.
[29] LI WJ， ZHAO ZG， LI X. The drought characteristics analysis in north China and its causes of formation[J]. Arid Meteorology2003， 21（4）： 1-5.
Key words Standardized precipitation index; Inland areas of Shandong Province; Drought; Characteristics
Drought is one of the most important natural meteorological disasters in China. It occurs frequently， has a long duration， and has a wide range of impacts. The occurrence of drought is likely to cause great agricultural production reductions， serious desertification， and lack of water resources， posing great impacts on normal agricultural production and people’s lives. Due to the influences of monsoon climate and landforms， Shandong Province is characterized by uneven distribution of water and heat and unbalanced precipitation， resulting in frequent occurrence of drought disasters. Therefore， the study and analysis on the drought characteristics of Shandong Province， especially the inland areas of Shandong Province， is of great significance to understand the occurrence and development trend of drought， timely adjust agricultural production activities and reduce the losses caused by drought.
As for the characteristics of drought， some scholars have proposed the Palmer Drought Index （PDSI）[1]， the Surface Water Supply Index （SWSI）[2] and the Standardized Precipitation Index （SPI）[3]. Among them， SPI is widely used because of its simple calculation， flexible time scale and better reflection of drought characteristics. At present， SPI has been more and more used to analyze the temporal and spatial variation characteristics and seasonal characteristic changes of droughts at home and abroad[4-19]， so as to better study the characteristics of drought evolution. Piccarreta[20] used SPI to analyze the drought in southern Italy from 1923 to 2000. Guttmann[21] compared SPI and PDI based on the precipitation data of the United States， and found that PDI spectral characteristics changed with the locations， presenting high uncertainty， while SPI did not， and SPI was much easier to use than PDI. SPI is one of the main bases of drought detection system in the United States[22]. In China， Zhang et al.[23] used SPI to analyze and study the drought and flood conditions and their occurrence frequency in the drought wetness episodes in the Pearl River Basin. Zhang et al.[23] analyzed the characteristics of drought evolution in Gansu Province based on SPI. Yuan et al.[4] analyzed SPI and Z index in China， and concluded that SPI could effectively analyze the drought situation in China. It could clearly reflect the changes of drought characteristics and had a good predictive effect on drought disasters， making it better than Z index. In this study， based on the monthly precipitation data （from the China Meteorological Data Service Center） of 5 meteorological stations in central and western Shandong Province from 1960 to 2013， the precipitation characteristics， SPI change characteristics and drought characteristics in the inland areas of Shandong Province from 1960 to 2013 were analyzed with SPI as the meteorological drought analysis index in the time scale from the aspect of meteorological drought， so as to provide a theoretical basis for drought monitoring， early warning and prevention of drought disasters.
Materials and Methods
Study area overview
The inland areas of Shandong Province have a warm temperate monsoon climate with concentrated precipitation， rain and heat falling in the same season， short spring and autumn， and long winter and summer. The annual average temperature is 11-14 ℃. The inland areas of Shandong Province have abundant light resources with the annual average sunshine time of 2 290-2 890 h， and the heat conditions can meet the needs of biannual cropping. The annual average precipitation is generally around 65-70 mm. The seasonal distribution of precipitation is very uneven. About 60%-70% of the annual precipitation falls in July-September. Moreover， it is easy to have flooding disasters in summer， and drought disasters in winter， spring and late autumn， which has a great impact on agricultural production.
Data and methods
Data acquisition
All precipitation data used in this paper was provided by the China Meteorological Data Service Center， including the month by month precipitation data of the 5 major meteorological stations in Jinan， Shen County， Yanzhou， Yiyuan and Taishan from 1960 to 2013. Statistically speaking， such a long time series was sufficient enough to obtain a more credible trend result.
SPI method
SPI is an indicator of the probability of occurrence of precipitation in a certain period of time， and it is suitable for drought detection and assessment relative to the local climatic conditions at the monthly or multimonthly time scales. The calculation of SPI is based on the assumption that the precipitation is a kind of skewed distribution. The Γ distribution probability is used to describe the change of precipitation， and the precipitation of the skewed probability distribution is normalized. Finally， the drought level is classified by the cumulative frequency distribution of standardized precipitation. The SPI calculation formula is as follows： SPI=St-（c2t+c1t）+c0d3t+d2t+d1t+1.0（1）
Where， t=ln1G（x） 2， and G（x） is the probability of precipitation distribution associated with the Γ function; x is the annual or seasonal precipitation sample; S is the positive and negative coefficients of probability density.
If G（x）>0.5， then S=1; if G（x）≤0.5， then S=-12. G（x） is calculated by the probability distribution integral formula of the Γ distribution function：
G（x）=1β γΓ（γ）0∫ x0xγ·1e-x/βdx， x>0（2）
Where， γ， β are the shape and scale parameters of the Γ distribution function; c0， c1， c2 and d1， d2， d3 are the calculation formulas for the simplified approximate solution formulas of the Γ distribution function converted to the cumulative frequency， and c0=2.515 517， c1=0.802 853， c2=0.010 328， d1=1.432 788， d2=0.189 269， d3=0.001 308.
According to the SPI values obtained by the above formulas， the SPI values for different drought levels were determined according to the frequency of occurrence of drought （Table 1）[24].
Linear regression method
On the basis of a large amount ofprecipitation data， the SPI value was calculated， and the mathematical statistics method was used to establish the regression function expression between the dependent variables and the independent variables， which could show the variation trends of SPI in the inland areas of Shandong Province.
Results and Analysis
Basic characteristics of precipitation in inland areas of Shandong Province from 1960 to 2013
The comparison on the mean， maximum， and minimum values of precipitation （Table 2） showed that the precipitation in Taishan area was the largest， while that in the western part of Shandong Province was relatively small. This was mainly due to the foehn effect. The Taishan side was on the windward slope， and the humid airflow was forced to rise when encountering the highlands like mountains. Thus， the temperature dropped， forming rainfall. On the other hand， the precipitation in western Shandong Province was little because it lied in the Huangpan Plain， where the shallow surface water evaporation effect was poor， and the groundwater storage was low， making it impossible to achieve precipitation cycle. At the same time， due to the influence of sea and land factors， it was impossible to carry out the precipitation cycle between land and sea. Therefore， the uneven distribution of precipitation was prone to result in uneven harvest situations in the study areas. In order to analyze the seasonal variation of drought， the variances of winter， spring， summer and autumn were calculated separately[29]. The results show that the variances of winter， spring， summer and autumn were 0.113， 0.402 676， 0.312 823 and 0.319 217， respectively in the inland areas of Shandong Province from 1960 to 2013. Therefore， the variation in spring was the biggest. The variation curve of SPI in each season was shown in Fig. 3.
The drought situation was obvious in spring in the inland areas of Shandong Province， and the SPI values showed no significant fluctuation. Besides， the spring drought occurred in the years of 1969， 1974， 1976， 1981， 1992 and 2012， in which the spring drought in 1968 was the most serious with the smallest SPI value， while SPI of the other years tended to be normal. The results showed that from the longterm trend， the spring droughts slowly decreased， and drought damage gradually reduced in the inland areas of Shandong Province. The spring precipitation tended to be stable， and the change trend was not obvious.
There was almost no drought in summer in the inland areas of Shandong Province， and the SPI values also tended to be stable. It was easy to have rain floods in summer， and the occurrence times of extreme weather somewhat increased in some periods. The results indicated that from the longterm run， there was no obvious drought in summer in the inland areas of Shandong Province.
On the other hand， the droughts in autumn were serious in the inland areas of Shandong Province. Severe droughts occurred in 1980， 1996 and 2009， especially after entering the 21st century， when the occurrence of drought became frequent. In addition， the variation of drought levels tended to be stable， and it was common to have light or moderate droughts， indicating that the autumn had high occurrence frequency of droughts in the inland areas of Shandong Province. The results showed that in the longterm trend， the autumn drought occurred frequently in the inland areas of Shandong Province， and the precipitation distribution in autumn was more uneven with more significant changes. Gradually， the autumn became the raindeficient season.
The drought conditions were extremely serious in winter the inland areas of Shandong Province， and droughts occurred throughout the whole winter. The droughts were very serious in the 1960s， 1980s， and early 21st century. The variation of drought levels was greater during the period 1974-1998 than in other time periods， and three severe a total of 3 heavy drought occurrences. It indicated that the frequency of drought occurrence was greater in winter than in any of the other 3 seasons in the inland areas of Shandong Province. Moreover， the longterm trend showed that winter drought was serious in the inland areas of Shandong Province， which was characterized for the extreme water shortage and significant changes. Conclusion
In this paper， SPI values of different time scales are calculated based on the month by month precipitation data from 5 inland areas in Shandong Province from 1960 to 2013 by using SPI as the drought assessment factor， and analysis has been made to the overall characteristics of precipitation， SPI interannual variation and seasonal distribution changes in these areas in the 54 years. Thus， the following conclusions are drawn.
（1） The drought characteristics change significantly in the inland areas of Shandong Province from 1960 to 2013. There are few occurrences of drought in spring and summer， while autumn and winter are the highincidence seasons of droughts. Moreover， it is easy to have moderate droughts in autumn， and severe droughts in winter.
（2） There is a general trend of drought in the inland areas of Shandong Province， and with the increase of time scale， SPI becomes much less random but more continuous. The occurrences of droughts were frequent in the 1980s and 1990s， and drought has also become the major natural disaster in the 21st century.
（3） In terms of interannual variations in drought， spring drought was the most severe in the 1960s， autumn drought was most severe in the 1980s， and winter drought was the most severe in the 1990s.
（4） As for the seasonal distribution of droughts， the area of spring drought gradually decreases， and there is almost no drought in summer. The drought area in autumn increases and gradually becomes more serious， and the drought in winter shows no significant variation. In general， the drought in autumn is the most serious throughout the whole province.
The analysis on the interannual and seasonal distribution characteristics of drought in the inland areas of Shandong Province shows that local agricultural production should pay full attention to the drought in spring and autumn， especially the occurrence of autumn drought. It should strengthen and improve the farmland irrigation system in the inland areas of Shandong Province， so as to improve the overall levels of drought prediction， detection and control in the areas. In this paper， the interannual and seasonal distribution changes are analyzed only from the SPI values at different time scales in the inland areas of Shandong Province from 1960 to 2013， but no consideration has been given to the continuity and variability of the drought across different seasons. In addition， as a complex meteorology， drought is a process of comprehensive action of the atmosphere， hydrosphere and biosphere. In this paper， the characteristics of drought are analyzed and studied only from the effects of precipitation， without considering the joint influences of the hydrological processes of the surface， atmospheric circulation and human activities， which should be the focus in further studies. References
[1] PALMER WC. Meteorological drought research paper[J]. 1965 （3）： 62-67.
[2] SHAFER BA， DEZMAN LE. Development of a Surface Water Supply Index （SWSI） to assess the severity of drought condition s in snow pack runoff areas Proceedings of the Western Snow Conference[N]. Colorado： Colorado State University， Fort Collins， 1982： 164-175.
[3] MCKEE TB， DOESKEN NJ， KLEIST J. The relation of drought frequency and duration to times cales1 Proceedings of the Eighth Conference on Applied Climatology[N]. Boston： Amer Meteorological Society， 1993： 179-1841.
[4] YUAN WP， ZHOU GS. Comparison between standardized precipitation index and Zindex in China[J]. Acta Phytocologica Sinica， 2004， 28（4）： 523-529.
[5] XUE DQ， WANG JG.. Characteristic analysis of change into drought in Shandong Province[J]. Journal of Natural Disasters， 2007， 16（3）： 60-65.
[6] MA JY， XU YL， PAN J， et al. Drought tendency based on standardized precipitation index （SPI） and relative moisture over northeast China from May to September during 1961-2009[J]. Journal of Meteorology and Environment， 2012， 28（3）： 90-95.
[7] SIMS AP， NIYOGI DS， RAMAN S. Adopting drought indices for estimating soil moisture： a North Carolina case study[J]. Geophysical Research Letters， 2002， 29（8）： 1183-1186.
[8] HUANG WH， YANG XG， LI MS， et al. Evolution characteristics of seasonal drought in the south of China during the past 58 years based on standardized precipitation index[J]. Transactions of the Chinese Society of Agricultural Engineering， 2010， 26（7）： 50-59.
[9] CHE SJ， LI CQ， SHEN SH. Analysis of droughtflood spatialtemporal characteristics based on standard precipitation index （SPI） in Hebei Province during 1965-2005[J]. Chinese Journal of Agrometeorology， 2010， 31（1）： 137-143.
[10] LI SJ， YUAN J， XIAO QH， et al. Characteristics of droughts and humidification based on standard precipitation index in Weifang City during the past 50 years[J]. Chinese Agricultural Science Bulletin， 2011， 7（26）： 293-298.
[11] LI WG， CHEN HL， ZHU NH， et al. Analysis of drought monitoring on Hainan Island from standardized precipitation index[J]. Chinese Journal of EcoAgriculture， 2009， 17（1）： 178-182.
[12] VICENTESERRANO SM. Spatial and temporal analysis of droughts in the Iberian Peninsula （1910-2000）[J]. Hydrological Sciences Journal， 2006， 51（1）： 83-97.
[13] GONG HQ， TAI QG. Analysis of rainfall characteristics in Linyi City[J]. Journal of Anhui Agricultural Sciences， 2008， 36（ 28）： 12356-12357. [14] YANG YQ. Statistical methods in agricultural meteorology[M]. Beijing： Meteorological Press， 1983， 153-160.
[15] SEIER RA， HAYES M， BRESSAN L. Using the standardized precipitation index for flood risk monitoring[J]. Int J of Climatology， 2002（22）： 1365-1376.
[16] PATEL NR， CHOPRA P， DADHWAL VK. Analyzing spatial patterns of meteorological drought using standardized precipitation index[J]. Meteorological Applications， 2007（14）： 329-336.
[17] EDWARDS DC， MCKEE TB. Characteristics of 20th Century Drought in the United States at Multiple Time Scales[D]. USA： Department of Atmospheric Science， Colorado State University， 1997： 18-21.
[18] ZHANG Q， ZOU XK， XIAO FJ. GB/T 20481-2006 Classification of meteorological drought[S]. Beijing： China Standard Press， 2006： 16-17.
[19] VICENTESERRANO SM， CUADRATPRATS JM. Trends in drought intensity and variability in the middle Ebro valley during the second half of the twentieth century[J]. Theoretical and Aplied Clinatology， 2007（88）： 247-258.
[20] PICCARRETA M， CAPOLOCONG D， BENZI F. Trend analysis of precipitation and drought in Basilcate from 1923 to 2000 within a southern Italy context[D]. Int. Climate， 2004， 24： 907-922.
[21] GUTTMAN NB. Comparing the Palmer drought index and the standardized precipitation index[J]. Journal for the American Water Resources Association， 1998， 34（1）： 113-121.
[22] SWOBODA D， HAYES M. The Drought Monitor Amer Meteor Soe[J]. 2002， 83： 1811-1190.
[23] ZHONG Q， XU CY， ZHANG ZY. Observed changes of drought wetness episodes in the Pearl River Basin[D]. China， 2009， 98： 89-99.
[24] General Administration of Quality Supervision， Inspection and Quarantine of the People's Republic of China， China National Standardization Administration. GB/T20481-2006 Classification of meteorological drought[S]. Beijing： China Standard Press， 2006.
[25] WEI CY， WANG JJ， ZHANG F， et al. Using expansion of empirical orthogonal function and geostatistics to analyze temporalspatial variability of precipitation[J]. Journal of Irrigation and Drainage， 2010， 29（4）： 105-109.
[26] EDWARD DC， MCKEE TB. Characteristics of 20th Century Drought in the United States at Multiple Times Scales[J]. USA： Department of Atmospheric Science， Colorado State University， 1997 ：18-21.
[27] HUANG WH， YANG XG， LI MS， et al. Evolution characteristics of seasonal drought in the south of China during the past 58 years based on standardized precipitation index[J]. Transactions of the Chinese Society of Agricultural Engineering， 2006， 26（7）： 50-59.
[28] CHEN J， LU DH. Division of monsoon climatic regions in China[J]. Journal of Beijing Forestry University， 1981： 34-41.
[29] LI WJ， ZHAO ZG， LI X. The drought characteristics analysis in north China and its causes of formation[J]. Arid Meteorology2003， 21（4）： 1-5.