Nanjing University of Information Science and Technology, Nanjing, Jiangsu Province, People’S Republic of China

分类号：______单位代码：______

密级：______学号：______

硕士学位论文

中国北方城市与沙漠地区气候响应对比

CLIMATE RESPONSE OVER SELECTED URBAN AND DESERT AREAS IN NORTHERN CHINA

申请人姓名：夏墨古 _ 指导教师：申双和教授 _ 合作教师： _ 专业名称：气象学 _ 研究方向：应用气候 _ 所在学院：国际教育学院 _

二〇一二年六月

CLIMATE RESPONSE OVER SELECTED URBAN AND DESERT AREAS IN NORTHERN CHINA

Dissertation Submitted to

NANJING UNIVERSITY OF INFORMATION SCIENCE & TECHNOLOGY

For the award of degree of

Master of Science in Meteorology

By:

Mugume Isaac

Dissertation Supervisor: Professor Shen Shuanghe

June 2012

独创性声明

本人声明所呈交的论文是我个人在导师指导下进行的研究工作及取得的研究成果。本论文除了文中特别加以标注和致谢的内容外，不包含其他人或其他机构已经发表或撰写过的研究成果，也不包含为获得南京信息工程大学或其他教育机构的学位或证书而使用过的材料。其他同志对本研究所做的贡献均已在论文中作了声明并表示谢意。

学位论文作者签名：______签字日期：______

关于论文使用授权的说明

南京信息工程大学、国家图书馆、中国学术期刊（光盘版）杂志社、中国科学技术信息研究所的《中国学位论文全文数据库》有权保留本人所送交学位论文的复印件和电子文档，可以采用影印、缩印或其他复制手段保存论文，并通过网络向社会提供信息服务。本人电子文档的内容和纸质论文的内容相一致。除在保密期内的保密论文外，允许论文被查阅和借阅，可以公布（包括刊登）论文的全部或部分内容。论文的公布（包括刊登）授权南京信息工程大学研究生部办理。

□公开 □保密（_____年 _____月） (保密的学位论文在解密后应遵守此协议)

学位论文作者签名：______签字日期：______

指导教师签名：______签字日期：______

DECLARATION

I, Mugume Isaac, declare that the work presented in this thesis is, to the best of my knowledge and belief my own research work. It is being submitted in partial fulfilment for the requirement for the award of Master of Science in Meteorology at the College of Applied Meteorology of Nanjing University of Information Science and Technology, Nanjing, Jiangsu Province, People’s Republic of China.

I further declare that this work has never been submitted in part or whole for any purpose to any academic institution or anywhere. The thesis is presented with the consent of my supervisor. The pieces of work by other authors, which were a source of reference and information, have been duly acknowledged by references to the authors.

Signature of candidate: Signature of supervisor :

AGREEMENT ON AUTHORIZED USE OF THESIS

Nanjing University of Information Science & Technology, China National Library, Chinese Academic Journal (CD) Magazine, China Degree Thesis Full Text Database of ISTIC have the right to keep the copies and electronic version of degree thesis, use methods of photocopy, reduction printing, etc. to keep the thesis, and provide information service to society by internet. The content of the electronic version thesis is in conformity with that of the paper thesis. The thesis is available for reference and borrowing and its full text or partial text can be published except those confidential theses during confidential period. Authorized publication of the thesis should be conducted by Graduate School, Nanjing University of Information Science & Technology.

口 Public 口 Confidential ( Year Month )

(Confidential thesis should obey this agreement when confidential period is ended)

Signature of Thesis Author: Date:

Signature of Supervisor: Date:

摘要

诸多研究已经证明全球气候变化背景下，区域气温与降水亦随时间推移发生变化，但关键问题在于不同地区、不同气象要素对气候变化的响应是保持一致，还是部分地区部分要素随时间呈现增加趋势，而另一部分地区另一部分要素呈现减少趋势。以此为切入点，研究选择 10 个位于北方城市市区气象站(参见 1.4.2 小节表 1.1)和同纬度 10 个沙漠腹地气象站(参见 1.4.2 小节表 1.2)1981 年至 2010 年逐日温度(最高气温、最低气温)和降水资料(参见 3.1 节)，基于 Mann-Kendall 趋势检验和回归分析方法探究中国北方城市与沙漠地区气温和降水变化趋势。

针对气温的研究结果表明：(1) 近 30 年来中国北方城市与沙漠地区春季气温日较差均呈减小趋势，但依据回归斜率，城市区域夏季气温日较差呈减小趋势(-0.140 oC/10 年)，沙漠地区夏季气温日较差呈增加趋势 (0.068 oC/10 年)。不同区域对比来看，城市地区气温日较差减小速率(春季：-0.307 oC/10 年)大于沙漠地区减小速率(春季：-0.023 oC/10 年)；而不同时间段对比来看，年均气温日较差减小速率要快于夏季、小于春季减小速率。(2) 沙漠地区日最高气温年际增速(0.510 oC/10 年)、春季增速 (0.540 oC/10 年)和夏季增速(0.550 oC/10 年)均大于城市区域增速(年际： 0.325 oC/10 年、春季：0.252 oC/10 年和夏季：0.389 oC/10 年)。城市和沙漠地区高温日数和极端高温事件也呈增加态势(详见 4.3.1.1 和 4.3.2.1 小节)。(3) 就日最低气温而言，中国北方城市与沙漠地区均呈升高态势，并且与日最低气温年际增速相比，沙漠地区春季和夏季日最低气温增速 (夏季：0.536 oC/10 年和春季 0.560 oC/10 年)均高于同纬度城市地区(夏季：0.529 oC/10 年和春季：0.560 oC/10 年)，而霜冻日数与极端低温事件均呈减少态势(详见 4.3.1.2 和 4.3.2.2 小节)。

针对降水的研究结果表明：(1) 中国北方沙漠地区降水总量总体呈现增加趋势，特别是春季降水总量增速更为显著。与此相比，同纬度城市区域降水总量仅在春季稍呈增加态势(M-K, z=0.095)且低于沙漠地区降水总量增速(M-K, z=0.160)，全年和夏季而言均表现为减少态势(M-K, z=- 0.058 和-0.133)。(2) 与降水总量类似，沙漠地区降水日数总体呈现增加态势，春季降水日数增速高于全年降水日数，夏季降水日数则呈减少态势。

同纬度城市区域降水日数仅在春季稍呈增加态势(M-K, z=0.041)且低于沙漠地区降水日数增速(M-K, z=0.101)，全年和夏季而言均表现为减少态势 (M-K, z=-0.064 和-0.131)。(3) 就极端降水事件而言，城市和沙漠地区春季极端降水事件基本一致，均呈增加态势，夏季极端事件变化则有差异，沙漠地区呈现增加态势而城市区域呈现减少趋势。

关键词：气候响应；Mann-Kendall趋势检验法；回归分析法；极端事件；中国北方城市与沙漠地区

ABSTRACT

Many studies carried out have shown evidence of regional temperature and precipitation variability along with global climate changes. A key issue is whether these variability follow trends that are uniform or others have increasing trends while others decreasing trend. And this study was carried out using Mann-Kendall trend test method and regression analysis method with the data observed by a total of 20stations to investigate the trends of temperature and precipitation in northern China. 10stations were selected from urban areas (section 1.4.2: table 1.1) and another 10stations were selected from desert areas of China (section 1.4.2: table 1.2). The data for temperature (maximum temperature, minimum temperature) and precipitation was obtained for these 20stations for the period 1981-2010 (section 3.1).

Results for temperature (sections 4.1.1 and 4.1.2), indicated that: the diurnal temperature range (DTR) trend was decreasing for both desert areas and urban cities in spring but decreasing for urban cities and increasing for desert areas in summer according to regression rates (urban cities: -0.140 oC/decade and deserts: 0.068 oC/decade). It was found out that DTR for the urban cities was decreasing at a faster rate (spring: -0.307 oC/decade) than that for the desert areas (spring: -0.023 oC/decade). The rate of decrease for DTR on annual scale was greater than for summer but smaller than for spring. Maximum temperature for the desert areas was increasing at a faster rate (annual: 0.510 oC/decade, spring: 0.540 oC/decade and summer: 0.550 oC/decade) than that (annual: 0.325 oC/decade, spring: 0.252 oC/decade and summer: 0.389 oC/decade) for the urban cities. The high temperature days and high temperature extremes for both the desert areas and urban cities were exhibiting an increasing trend (section 4.3.1.1 and 4.3.2.1). The minimum temperature was also increasing for both desert areas and urban cities. It was also found that minimum temperature for the desert areas was increasing at a faster rate than that for the urban cities for summer and spring seasons (deserts: 0.536 oC/decade and 0.560 oC/decade and urban cities: 0.529 oC/decade and 0.560 oC/decade) compared to annually (deserts: 0.520 oC/decade, urban cities: 0.566 oC/decade). The frost days and low temperature extremes for both the desert areas and urban cities were exhibiting a decreasing trend (section 4.3.1.2 and 4.3.2.2).

III

For precipitation, the results obtained in this study indicated that: the rainfall amount was increasing for desert areas at a rate greater than that of the selected urban cities with increase in spring season more significant than the annual scale and summer seasons. For the urban cities, the rainfall amount was even decreasing on annual scale and in summer season (M-K, z=-0.058 and -0.133) with a slight increase in spring season (M-K, z=0.095) which was still less than the rate for the desert areas (M-K, z=0.160). The rain days are increasing for desert areas at a rate greater than that of the selected urban cities. The increase was more significant in the spring season than the annual scale and the summer season was also witnessing decreasing rain days. For the urban cities, the rain days were even decreasing on annual scale and in summer season (M-K, z=- 0.064 and -0.131) with a slight increase in the spring (M-K, z=0.041) which was still less than the rate for the desert areas (M-K, z=0.101). For precipitation extremes, both the desert areas and urban cities were exhibiting an increasing trend of spring precipitation extremes and that of summer precipitation extremes in deserts but decreasing trend of summer precipitation extremes in urban cities (section 4.3.1.3 and 4.3.2.3).

Key words: climate response; Mann-Kendall trend test method; regression analysis method; extreme events; urban and desert areas in northern China

TABLE OF CONTENT

Page 独创性声明关于论文使用授权的说明 DECLARATION AGREEMENT ON AUTHORIZED USE OF THESIS 摘要 I ABSTRACT III LIST OF EQUATIONS, FIGURES AND TABLES VIII ABBREVIATIONS AND ACRONYMS XII

CHAPTER ONE 1 1.0 INTRODUCTION 1 1.1 BACKGROUND TO THE STUDY 1 1.2 AIM FOR THE STUDY 6 1.2.1 General Aim 6 1.2.2 Objectives of the study 6 1.3 JUSTIFICATION FOR THE STUDY 7 1.4 SCOPE OF STUDY 8 1.4.1 Climate Parameters 8 1.4.2 Geographical area 8

CHAPTER TWO 11 2.0 LITERATURE REVIEW 11 2.1 CLIMATE CHANGE AND CLIMATE VARIABILITY 11 2.2 TEMPERATURE 12 2.3 PRECIPITATION 15 2.4 DIURNAL TEMPERATURE RANGE AND PRECIPITATION 17 2.5 REMAINING PROBLEM: COMMENT ON LITERATURE REVIEW 17

CHAPTER THREE 18 3.0 DATA AND METHODOLOGY 18 3.1 DATA USED IN THE STUDY 18

3.1.1 Temperature data 18 3.1.2 Precipitation data 19 3.2 METHODS USED IN THE STUDY 19 3.2.1 The relative deficit or surplus (relative anomalies) 19 3.2.2 Standard deviation and coefficient of variation 20 3.2.3 The Mann-Kendall trend test 21 3.2.4 The Beard formula for climate extremes 22

CHAPTER FOUR 24 4.0 RESULTS AND DISCUSSION 24 4.1 TEMPERATURE RESPONSE 24 4.1.1 Over desert areas 24 4.1.1.1 Annual temperature trends 24 4.1.1.2 Spring temperature trends 27 4.1.1.3 Summer temperature trends 31 4.1.2 Over selected cities 34 4.1.2.1 Annual temperature trends 34 4.1.2.2 Spring temperature trends 37 4.1.2.3 Summer temperature trends 40 4.2 PRECIPITATION RESPONSE 43 4.2.1 Over desert areas 44 4.2.1.1 Annual precipitation trends 44 4.2.1.2 Spring (MAM) precipitation trends 46 4.2.1.3 Summer precipitation trends 49 4.2.2 Over selected cities 51 4.2.2.1 Annual precipitation trends 51 4.2.2.2 Spring (MAM) precipitation trends 54 4.2.2.3 Summer precipitation trends for selected urban cities 56 4.3 CLIMATE EXTREMES 58 4.3.1 Over desert areas 58 4.3.1.1 High temperature days and high temperature extremes 58 4.3.1.2 Frost days and low temperature extremes 59 4.3.1.3 Precipitation extremes 60 4.3.2 Over selected cities 61 4.3.2.1 High temperature days and high temperature extremes 61

4.3.2.2 Frost days and low temperature extremes 62 4.3.2.3 Precipitation extremes 63

CHAPTER FIVE 66 5.0 CONCLUSION AND RECOMMENDATIONS 66 5.1 TEMPERATURE RESPONSE 66 5.1.1 DTR trends 66 5.1.2 Maximum temperature trends 67 5.1.3 Minimum temperature trends 67 5.2 PRECIPITATION RESPONSE 68 5.2.1 Trends of rainfall amounts 68 5.2.2 Trends of rain days 69 5.3 CLIMATE EXTREMES 70 5.3.1 High temperature days and high temperature extremes 70 5.3.2 Frost days and low temperature extremes 70 5.3.3 Precipitation extremes 71 5.4 SUMMARY AND RECOMMENDATIONS 72

REFERENCES AND BIBLIOGRAPHY 74 APPENDICES 80 ACKNOWLEDGEMENT 90

VII

LIST OF EQUATIONS, FIGURES AND TABLES

Page

Equations Equation for DTR, equation (3.1) 18 Equation for relative deficit (or surplus), equation (3.2) 19 Equation for average, equation (3.3) 20 Equation for standard deviation, equation (3.4) 20 Equation for coefficient of variation, equation (3.5) 21 Equations for Mann-Kendall: Equation (3.6) 21 Equation (3.7) 21 Equation (3.8) 21 Equation (3.9) 22 Beard formula for climate extremes: Equation (3.10) 22

Figures Fig. 1.1: Map of China showing the study locations 9 Fig. 2.1: Global temperature trends 14 Fig. 2.2: Global precipitation climatology 16 Fig. 4.1: Annual DTR relative anomalies for selected desert areas. 26 Fig. 4.2: Annual maximum temperature relative anomalies for selected desert areas. 26 Fig. 4.3: Annual minimum temperature relative anomalies for selected desert areas. 27 Fig. 4.4: Spring DTR relative anomalies for selected desert areas. 29 Fig. 4.5: Spring maximum temperature relative anomalies for selected desert areas. 30 Fig. 4.6: Spring minimum temperature relative anomalies for selected desert areas. 30 Fig. 4.7: Summer DTR relative anomalies for selected desert areas. 32

VIII

Fig. 4.8: Summer maximum temperature relative anomalies for selected desert areas. 33 Fig. 4.9: Summer minimum temperature relative anomalies for selected desert areas. 34 Fig. 4.10: Annual DTR relative anomalies for selected cities. 36 Fig. 4.11: Annual maximum temperature relative anomalies for selected cities. 36 Fig. 4.12: Annual minimum temperature anomalies for selected urban areas. 37 Fig. 4.13: Spring DTR relative anomalies for selected cities. 39 Fig. 4.14: Spring maximum temperature relative anomalies for selected cities. 39 Fig. 4.15: Spring minimum temperature relative anomalies for selected cities. 40 Fig. 4.16: Summer DTR relative anomalies for selected cities. 42 Fig. 4.17: Summer maximum temperature relative anomalies for selected cities. 43 Fig. 4.18: Summer minimum temperature relative anomalies for selected cities. 43 Fig. 4.19: Annual trend of rain days relative anomalies for desert regions. 45 Fig. 4.20: Annual trend of rainfall amount relative anomalies for desert regions. 46 Fig. 4.21: Spring trend of rain days relative anomalies for desert regions. 48 Fig. 4.22: Spring trend of rainfall amount relative anomalies for desert regions. 48 Fig. 4.23: Summer trend of rain days relative anomalies for desert regions. 50 Fig. 4.24: Summer trend of rainfall amount relative anomalies for desert regions. 51 Fig. 4.25: Annual trend of rain days relative anomalies for selected cities 52 Fig. 4.26: Annual trend of rainfall amount relative anomalies for selected cities. 53 Fig. 4.27: Spring trend of rain days relative anomalies for selected cities. 55

Fig. 4.28: Spring trend of rainfall amount relative anomalies for selected cities. 55 Fig. 4.29: Summer trend of rain days relative anomalies for selected cities. 57 Fig. 4.30: Summer trend of rainfall amount relative anomalies for selected cities. 57

Tables Table 1.1: Selected city stations. 9 Table 1.2: Selected desert stations. 10 Table 4.1: M-K trend (z) results for annual temperature – Desert 25 Table 4.2: M-K (z) results for spring temperature – Desert 28 Table 4.3: M-K trend (z) results for summer temperature – Desert 31 Table 4.4: M-K trend (z) results for annual temperature – Cities 35 Table 4.5: M-K trend (z) results for spring temperature – Cities 38 Table 4.6: M-K trend (z) results for summer temperature – Cities 41 Table 4.7: M-K trend (z) results for annual precipitation – Deserts 44 Table 4.8: M-K trend (z) results for spring precipitation – Deserts 47 Table 4.9: M-K trend (z) results for summer precipitation – Deserts 50 Table 4.10: M-K trend (z) results for annual precipitation – Cities 52 Table 4.11: M-K trend (z) results for spring precipitation – Cities 54 Table 4.12: M-K trend (z) results for summer precipitation – Cities 56 Table 4.13: M-K trend (z) results for high temperature days and extremes – deserts 59 Table 4.14: M-K trend (z) results for frost days and low temperature extremes –deserts 60 Table 4.15: M-K trend (z) results for precipitation extremes – deserts 61 Table 4.16: M-K trend (z) results for high temperature days and extremes – cities 62 Table 4.17: M-K trend (z) results for frost days and low temperature extremes –cities 63

Table 4.18: M-K trend (z) results for precipitation extremes – cities 64 Table 5.1: Average results for DTR using Mann-Kendall trend (z) test 66 Table 5.2: DTR – Regression 66 Table 5.3: Average results for maximum temperature using M-K trend (z) test. 67 Table 5.4: Maximum temperature trend – regression. 67 Table 5.5: Average results for minimum temperature using M-K trend (z) test. 68 Table 5.6: Minimum temperature trend – regression. 68 Table 5.7: Average results for rainfall trend using M-K trend (z) test. 68 Table 5.8: Average results for rain days trend using M-K trend (z) test. 69 Table 5.9: Average results for high temperature days and high temperature extremes trend using M-K trend (z) test. 70 Table 5.10: Average results for frost days and low temperature extremes trend using M-K trend (z) test. 71 Table 5.11: Average results for Precipitation extremes trend using M-K trend (z) test. 71

ABBREVIATIONS AND ACRONYMS

AMS : American Meteorological Society Avg : Average ENSO : El Nino Southern Oscillation Fig. : Figure GHGs : Greenhouse gases High Temp. Ext: High temperature extremes IPCC : Intergovernmental Panel on Climate Change JJA : Summer season (June, July, August) L. Temp. Ext : Low temperature extremes MAM : Spring season (March, April, May) Max. Temp : Maximum temperature (in degrees Celsius, oC) Min. Temp : Minimum temperature (in degrees Celsius, oC) M-K : Mann-Kendall N/A : Not applicable

NOx : Oxides of Nitrogen

O3 : Ozone PAN : Peroxyacytylnitrate prcp : Precipitation (in millimetres, mm) R.Days : Rainfall Days (or Rainy days) RF : Radiative Forcing SST : Sea Surface Temperature (in degrees Celsius, oC) o Tmax : Maximum temperature (in degrees Celsius, C) o Tmin : Minimum temperature (in degrees Celsius, C) TOT-prcp : Total annual precipitation (in millimetres) z : Statistic to represent Mann-Kendall none parametric test for trend.

XII

CHAPTER ONE INTRODUCTION CHAPTER ONE

1.0 INTRODUCTION

This chapter presents the background and idea behind the study. Section 1.1 gives the background to the study including defining key terms in the study. Section 1.2 has listed the broad aim of the study as well as the specific objectives of the study. Section 1.3 gives the problem statement, thereby presenting the motivation (justification) behind the study. Section 1.4 has the scope of study, presenting the data scope that has been used as well as the geographical location of the study.

1.1 Background to the study

Rao (2008) has defined weather as the physical state of atmosphere at a given point in time and at a given location. Weather 1 , which is the state of the atmosphere at some place and time described in terms of such variables as temperature, cloudiness, precipitation and so on, can have positive or negative impact on economic activity thereby affecting directly or even indirectly production and consumption (Lazo et. al., 2011; Morss et. al., 2008). Climate2 on the other hand is the statistical description in terms of mean and variability of relevant quantities over a period of time [sic]. In this study, the trends of two important aspects of weather and climate are being studied which are temperature and precipitation.

The Temperature of a body is its thermal state in reference to its ability of communicating heat to other bodies (Maxwell & Pesic, 2001; Srivastara, 2008) and is registered (or measured) by using a thermometer. Srivastara (2008) further explained that; Meteorologists are interested in the temperature of air near the ground. A set of thermometers, which measures air temperature, are put in a Stevenson screen at a height in the range of 1.2m and 2m from the ground (Kimei & Khabongo, 2008; Linacre, 1992) and air temperature measured at this height is considered to be representative of a wide area. The daily variation in air temperature near the earth’s surface is controlled mainly by incoming solar energy and outgoing long-wave surface radiation (Ahrens, 2008). Precipitation,

1 http://www.ametsoc.org/amsedu/wes/glossary.html#W 2 http://www.wmo.int/pages/themes/climate/understanding_climate.php

DISSERTATION SUBMITTED TO NANJING UNIVERSITY OF INFORMATION SCIENCE & TECHNOLOGY in millimetres or inches, is the amount of liquid water depth of the substance that has fallen at a given point over a specified period of time (Michaelides et al., 2009). Precipitation largely varies in spatial and temporal (Feuerbacher & Stoewer, 2006; Michaelides et al., 2009; Wang & Cho, 1997) extent and there is significant marked difference between seasons as well as differences in precipitation regimes over the globe.

The importance of temperature, precipitation and global radiation have been explained by Baudoin and Zabeltitz (1999) as key parameters of classifying climatological zones, for instance: temperate, humid, desert and semi-desert. Precipitation on one hand supports agriculture and also influences other climatic phenomenon like droughts and floods. Hanif (2005) explained that: even though the seasonal distribution of precipitation is an important attribute to climate of a region, diurnal cycle of precipitation influences the precipitation effectiveness in agriculture. The amount of rainfall received is also important. This is so because: if precipitation is in excess, there is a likelihood of floods (Watson et al., 2010; Olivier et al., 2007). In regard to the relationship between precipitation and floods, Stefan et al. (2002) showed that precipitation was significantly correlated to floods and if it is insufficient, there is a likelihood of droughts (Harris and Kay, 1994 and Maslin, 2009). All these extreme events (floods and droughts) have impact on human kind and development and sometimes lead to death (Maslin, 2009).

As already noted, temperature and precipitation are among the elements of weather (Maloney, 2008). Rao (2008) further listed other elements of weather which include: wind (speed, direction and force), clouds (amount, type and height of cloud base from the ground), sunshine, pressure and humidity. Saucier (2003) explained that: these weather elements vary from day to day and time to time of the day (diurnal variation). It is important to agree that meteorological parameters do not vary in isolation. For instance, Bloomer et al. (2010) in their study of changes in seasonal and diurnal cycles of ozone and temperature over eastern US found that: both high temperatures and high pollution concentrations being associated with synoptic high pressure systems and sunny (with fast photolysis rates) stagnant conditions. On large scale: however, over a large area and over a long time, these weather elements tend to have an average. This average defines the climate of the area which was defined by Rao (2008) as the long-term regime

CHAPTER ONE INTRODUCTION of the atmospheric variables or the composite of the day-to-day values of the weather elements over a long period of time of a given place of area.

While weather elements vary from day to day or even time to time (Saucier, 2003), climate is also not constant over a long time (Murphey, 1982). It varies from season to season and year to year. Murphey (1982) also noted that climate fluctuations have influenced people’s lives, even though they have tended to balance each other out. One of the evident fluctuations in terms of seasons, that is; from season to season, was noted by Anderson (1975) who observed that climatic trend may be different in different seasons. Studies and meteorological data have shown that even within season, or annual data, there have been fluctuations (Kubota et al., 2005; Waylen & Caviedes, 1990). It is important to note that there are climatic fluctuations such as El Nino Southern Oscillation (ENSO), and that there is a difference between climatic variability and climate change. According to Bloomer et al. (2010), these changes in climate may manifest themselves as changes not just in the mean state, but also in variability or in diurnal or seasonal cycles. Over the years, there have been considerable studies about climate change (Stone et al., 2001) and due to its possible effects; climate change has been a significant topic of research (Aksu et al., 2010). There is an established body, Intergovernmental Panel on Climate Change (IPCC) to study and advise issues regarding climate change. IPCC is therefore the authority on issues regarding climate change. Regarding possible causes of climate change, Baas and Selvaraju (2007) have attributed climate change to both natural variability and human activities.

The previous studies conducted about weather parameters have demonstrated that; during the past decades, precipitation, temperature, and other climatic parameters as well as vegetation cover have changed significantly (Wang et al., 2011) and other studies have linked economic development and urbanization (Schmal, 1981). These changes appear to influence each other directly or indirectly. Baas and Selvaraju (2007) have explained that; a change in one weather element can produce changes in regional climate. They have considered, for example that; if average temperature increases significantly, it can affect the amount of cloudiness as well as the type and amount of precipitation that occurs. Sailor (1993) has also explained the effect of urbanization on landscape, notably; lowering reflectivity to solar radiation, surface moisture availability, modifying vegetation cover as well

DISSERTATION SUBMITTED TO NANJING UNIVERSITY OF INFORMATION SCIENCE & TECHNOLOGY as anthropogenic heat release. He further explained the said effect combine and cause temperature rise.

The IPCC (4th assessment) report has noted that; the dominant aspect of land cover change since 1750 has been due to deforestation in temperate regions. The overall effect of anthropogenic land cover change on global temperature will depend largely on the relative importance of increased surface albedo in winter and spring (exerting a cooling) and reduced evaporation in summer and in the tropics (exerting a warming). IPCC further noted that; estimates of global temperature responses from past deforestation vary from 0.01°C to 0.25°C. IPCC further explains that; if cooling by increased surface albedo dominates, then the historical effect of land cover change may still be adequately represented by Radiative Forcing (RF). With tropical deforestation becoming more significant in recent decades, warming due to reduced evaporation may become more significant globally than increased surface albedo.

Anthropogenic emissions, deforestation and urbanization and other surface characteristics changes have been accused to lead to temperature variations (Sailor, 1993) including climate change (Baas & Selvaraju, 2007). Sailor (1993) for instance explains that, urbanization results in a landscape with significantly modified surface characteristics which lead to lower values of reflectivity to solar radiation, surface moisture availability, and vegetative cover, along with the higher values of anthropogenic heat release and surface roughness. These combine to result in a higher air temperatures in urban areas compared to the rural areas. We would also assume therefore the trends of temperature and precipitation for desert areas not to follow a similar trend like the urban areas, or the rate to be lower than that observed for the case of urban areas in line with Sailor (1993). It has also been noted that; development is directly correlated with urbanization (Watson et al. 1998; Ahmed, 2007). Urbanization is directly correlated with decrease in surface greenness and hence changes in surface albedo (Sailor, 1993). These changes affect receipt of radiation on the earth and Influence air (surface) temperatures. Urbanisation and industrialisation have led to increased pollution, in terms of greenhouse gases (GHGs) which has also influenced temperature (Rooij, 2006). When it comes to China, in recent years (say from about; 1978) China has experienced massive industrial growth and development as well as increased urbanisation (Zhou et al. 2004). In China, cities have come up, which have

CHAPTER ONE INTRODUCTION influenced the local weather leading to urban heat island effect (Sailor, 1993). As for this study, the cities that have been considered include; Beijing, Changchun, Dalian, Datong, Harbin, Hohhot, Shenyang, Taiyuan, Tangshan and Tianjin. Details of these cities, there location, approximate geographical area and population have been presented in the table (1.1) under sub-section (1.4.2).

Further still with deserts and semi-deserts, these are extensive areas occupying about 1/3rd of the global land surface (Laity, 2008). The deserts are considered to be warm and can be divided into classes on basis of their air temperature (as; hot, temperate or coastal). In simple terms, drier areas of the earth are simply called deserts or semi-deserts. According to US department of Interior and US Geological Survey, in their scientific report, they noted that temperatures are mild in winter and hot in summer and that diurnal temperature range could be about 30oF [sic]. Kusky (2009) have illustrated that climate change is causing many of the deserts of the world to expand.

Although there could be minor differences in characteristics of deserts, Laity (2008) has noted the following general characteristic for deserts; (1) They have high summer temperature, (2) An excess of potential evaporation over precipitation as a result of high temperatures, winds and clear skies, (3) High variability of precipitation in terms of totals, distribution and intensity, (4) More prominent role of wind, (5) Clear skies prevailing most of the time and (6) Low humidity. In this study, desert areas of China that have been considered, include; Alxa Zuoqi, Bayan MOD, Da-Qaidam, Guaizihu, Hami, Hoboskar, Linhe, Qiemo, Ruoqiang and Turpan. Details of these deserts and their location have been presented in table (1.2) under sub-section (1.4.2)

Regarding the scope; the scope of climate change is wide. This study has committed itself to studying the trends of precipitation and temperature. For precipitation; trend analyses of rainfall deal with variations of amount of rainfall with time (Lau & Mink, 2006). They further noted that, these studies can be conducted for wet and dry seasons, calendar months, as well as year-by-year and even hour-to-hour. Regarding temperatures, Watson et al. (1998) have noted that climate change is likely to increase the frequency of very hot days. Increased temperatures have effects worth to worry about. For instance, Watson et al. (1998) have further given evidence that, analysis of concurrent meteorological and

DISSERTATION SUBMITTED TO NANJING UNIVERSITY OF INFORMATION SCIENCE & TECHNOLOGY mortality data in cities in the Middle East indicate that overall death rates rise during heat waves. Also, increased temperature threatens water reservoirs systems, resulting to increased incidence and severity of heat-waves, droughts (Oerlemans, 2001) and shrinking of glaciers ice caps, mountain glaciers, and permafrost regions of the world (Kusky, 2009). All these present cascading effects to humanity. In fact, Laity (2008) has explained that, increased temperatures will lead to loss of glaciers due to melting of water from glaciers thereby, impacting on human fauna and flora communities through interior deserts in Asia. Buda and DeWalle (2002) have warned that while changes in temperature and precipitation due to climate change will have adverse effects on natural and social resources and human health, atmospheric processes such as acidic deposition will also be affected.

1.2 Aim for the study

1.2.1 General Aim

The general aim of the study was to: investigate the climate response in terms of the trends in temperature and precipitation as well as their extremes over selected urban and desert areas in China over a period of 30years from 1981 to 2010.

1.2.2 Objectives of the study

The objectives for the study, over the study period (1981 – 2010) were to: i) Compare temperature (maximum, minimum and diurnal) trends over selected urban areas and desert region areas in China with respect to annual, spring season (March to May) and summer (June to August). ii) Analyse trends of precipitation (including rainy days) over the study areas for the study period with respect to annual, spring season (March to May) and summer (June to August) iii) Examine the trend of climate extremes (the number of days with maximum temperatures over 35.0oC, the number of days with minimum temperatures below 0oC, the number of days with maximum temperatures over the 95th percentiles of daily maximum temperatures, the number of days with minimum temperatures below the 5th percentiles of daily

CHAPTER ONE INTRODUCTION

minimum temperatures, and the number of days with daily precipitation over the 99th percentiles of daily precipitation).

1.3 Justification for the study

There have been enormous studies about temperature trends. If these temperature trends increase at the evidenced rate, Laity (2008) has exposed a threat that: the ability of some ecosystem species like birds may not psychologically respond to increased temperature. They may in fact migrate as is usually the case during season to season (Cox, 2010). The other case is change in the occurrence of extreme weather events. For example, Baas and Slevaraju (2007) considered the case of Bangladesh and explained that: small changes in average conditions can have big influence on extremes such as droughts. The effects related to increased temperature, may therefore include: increased incidence of droughts (Bass & Slevaraju, 2007; Kusky, 2009), cloudiness (Cox, 2010), desertification (Kusky, 2009), heat-wave (Maslin, 2009) and forest fires among others.

As for precipitation, increased precipitation has disastrous effects too. For instance, it may directly lead to floods for low lying areas with poor drainage (Watson et al. 2010; Olivier et al. 2007), displacement of people, crop failure due to water logging (Ninno et al. 2001) or even washing away (erosion) of soil and planted crops (as is usually the case for hilly areas), disruption in transport and destroying infrastructure including buildings (Olivier et al. 2007), roads and bridges. On the other hand, suppressed precipitation leads to droughts (Maslin, 2009) which have a lethal effect on agriculture especially crop yields.

This study has attempted to study the climate response regarding the characteristics of temperature and precipitation trends by comparing urban areas and desert areas of China. The trend of extreme climatic events (high temperature days, frost days, high temperature extremes, low temperature extremes, and precipitation extremes) has also been studied. In this study, the high temperature days are the number of days with maximum temperatures over 35.0oC. The frost days are the number of days with minimum temperatures below 0oC. The high temperature extreme days are the number of days with maximum temperatures over the 95th percentiles of daily maximum temperatures. The low temperature

DISSERTATION SUBMITTED TO NANJING UNIVERSITY OF INFORMATION SCIENCE & TECHNOLOGY extreme days are the number of days with minimum temperatures below the 5th percentiles of daily minimum temperatures. The precipitation extreme days are the number of days with daily precipitation over the 99th percentiles of daily precipitation. A key question under study was whether these trends are the same for urban areas and desert areas, with keen interest to find out which trend is worsening so as to present a foundation for further research.

1.4 Scope of study

1.4.1 Climate Parameters

The climatic parameters studied are temperature and precipitation. The study has been in regard of the climate response of temperature and precipitation trends over the study region and study period. The study period has been chosen from 1981 to 2010, inclusive. The temperature has further been studied by dividing it into: maximum temperature, minimum temperature and diurnal temperature range (DTR). For precipitation, this study has considered: precipitation amount and rain days. In terms of time scale, the study has been done in terms of: annual, spring season and summer season. The comparisons as well as trends for the climate variables under study have been done in that sense of the time scale.

1.4.2 Geographical area

The study area has been confined to selected areas (selected cities and selected areas in the desert region) of the northern region of People’s Republic of China (PRC), or simply, North China. The figure (Fig. 1.1) below represents the map of China and the study areas.

CHAPTER ONE INTRODUCTION

Fig. 1.1: Map of China showing the study locations.

The cities that have been considered for the study include; Beijing, Changchun, Dalian, Datong, Harbin, Hohhot, Shenyang, Taiyuan, Tangshan and Tianjin. Details of location, urban population and approximate urban area of the cities under study have been presented in the table (1.1) below

Table: 1.1 City Stations3 Urban Urban Area Station ID LAT LONG Population (Sq. mile) (as by 2010) Beijing 54511 39.93N 116.28E 6,487 19,612,368 Changchun 54161 43.86N 125.33E 1,808 3,341,700 Dalian 54662 38.90N 121.63E 4,855 3,578,000 Datong 53487 40.10N 113.33E 803 1,570,035 Harbin 50953 45.75N 126.77E 807 4,517,549 Hohhot 53463 40.82N 111.68E 833 1,980,774 Shenyang 54342 41.77N 123.43E 1,338 5,743,718 Taiyuan 53772 37.78N 112.55E 564 3,212,500 Tangshan 54534 39.65N 118.18E 1,388 3,163,152 Tianjin 54527 39.10N 117.17E 66 4,342,770

3 Details about geographical area and population of the cities have been obtained from http://en.wikipedia.org

DISSERTATION SUBMITTED TO NANJING UNIVERSITY OF INFORMATION SCIENCE & TECHNOLOGY According to MacKinnon and Phillips (2006), desert areas in China accounts for almost 30% of total land area of China. The major deserts in China are located in Xinjiang province (Longjun and Xiaohui, 2010) and others in Inner Mongolia. Ten (10) stations from these deserts have been considered for this study and they include; Alxa Zuoqi, Bayan MOD, Da-Qaidam, Guaizihu, Hami, Hoboksar, Linhe, Qiemo, Ruoqiang and Turpan. Details of these deserts are given in table (1.2) below

Table: 1.2 Desert Stations Station Name ID LAT LONG Alxa Zuoqi 53602 38.83N 105.53E Bayan MOD 52495 40.75N 104.50E Da-Qaidam 52713 37.85N 095.62E Guaizihu 52378 41.37N 102.37E Hami 52203 42.82N 093.52E Hoboksar 51156 46.78N 085.72E Linhe 53513 40.77N 107.40E Qiemo 51855 38.15N 085.55E Ruoqiang 51777 39.03N 088.17E Turpan 51573 42.93N 089.20E

CHAPTER TWO LITERATURE REVIEW CHAPTER TWO

2.0 LITERATURE REVIEW

In this chapter, detailed discussions about concepts in this study have been presented as well as previous research that relates to the study. The first section (2.1) gives an overview of climate change and climate variability, the second section (2.2) presents details of studies conducted about temperature, section (2.3) has considered precipitation and section (2.4) handles relationship between DTR and precipitation. The last section, section 2.5 explains the remaining problem and gives a concise comment on Literature Review in regard to study concept.

2.1 Climate change and climate variability

Kusky (2007) has defined variations in average weather at different times of the year as seasons. When it comes to climate; climate fluctuations have attracted people’s attention and have linked such fluctuations to climate change (Murphey, 1982). It is important to have a thorough understanding of climate variability and trends. Climate variability is said to occur when the climatic parameter of a region has varied from its long-term mean (Baas and Selvaraju, 2007). Climate variability should not be confused with climate change. Baas and Selvaraju (2007) have also defined climate change as any change in climate over time, whether due to natural or anthropogenic forces.

Studies conducted regarding climate change link the change to basically two causes, namely: anthropogenic (human caused) and natural causes. Considering the anthropogenic causes, Laity (2008) for instance explained that, by changing the properties and scale of naturally occurring land-surface elements, human influences can significantly affect weather, climate and the chemistry and aerosol loading of the atmosphere. Qian et. al. (2011) advises that: climate change should not only be reflected or implied in the changes in annual means of climate variables but also in the changes in their annual cycles.

As for emission of greenhouse gases (GHGs), researches (Akinremi et al., 1999) have attributed the recent increasing trends in greenhouse gases to human activity. Zhang et. al. (2011) have explained this issue further that; global warming being

DISSERTATION SUBMITTED TO NANJING UNIVERSITY OF INFORMATION SCIENCE & TECHNOLOGY caused by human-induced emission of greenhouse gases is accelerating the global hydrological cycle. They further note that: the accelerated hydrological cycle is altering the spatial–temporal patterns of precipitation resulting in increased occurrences of precipitation extremes and in turn increased occurrences of floods and droughts in many regions of the world. The other case has been presented by Laity (2008), who has argued that a doubling of Carbon dioxide (CO2) in global atmosphere is likely to increase the area of desert land by about 17%. The increasing trend of GHGs is in line with increasing temperature trends, indicating a strong relationship. Urban areas, with their high population, emission from automobile and industrial emission have put urban areas at spot regarding temperature trends and the urban heat effect (Sailor, 1993).

2.2 Temperature

A lot of studies about climate trends have been carried by a couple of scholars over the years. For instance, Harger (1995) one of the scholars, studied the increasing annual trend in air-temperature exhibited by the mean monthly values over the period 1866-1993, for the Jakarta and the Semarang data. The studies conducted by Harger (1995) concentrated on temperature trends for ENSO and non ENSO years of Jakarta and Semarang. Results showed that; the coldest months for ENSO years became slightly lower than those for non ENSO years. Harger’s study revealed that the air temperature trend taken together is 1.64°C (0.0132°C per year from 25.771to 27.409°C). Harger (1995) suggested that the 1.65 ° difference between 1866 and 1991 could have been as a result of: (1) urban heat-island effect, (2) effect of deforestation, (3) effect of secular micro-climate shift and (4) influence of general global warming with particular reference to the tropics

The earth is not homogeneous; it has both land and water bodies as well as different altitudes with highlands, valleys and flat areas 4 . In addition to the difference in surface characteristics, Barry and Chorley (1998) have explained that different parts of the earth’s surface receive different amounts of solar

4 Details about the nature of the earth can be accessed from: http://en.wikipedia.org/wiki/Earth

CHAPTER TWO LITERATURE REVIEW radiation. The other factor is the time of the year, where more radiation is received in summer months than winter. Barry and Chorley (1998) also points out the importance of latitude in that, it contributes to the duration of daylight and the distance travelled through the atmosphere by the oblique rays of the sun. The receipt of solar radiation on the earth surface contributes to surface temperature among other factors (Wang et al., 2011).

According to Roy et al. (2009), one of the processes that could influence local temperature is the high degree of urbanization. This is in agreement with the fourth assessment report of the IPCC which provided evidence of the connection between urbanization-related human activities and the increased incidence of hot days, hot nights, and heat waves in the recent years. Many studies about temperature have been carried out (Bloomer et al., 2010; Buda et al., 2002; Harger, 1995; Maslin, 2009; Roy et al., 2009 and Sailor, 1993). These studies have also considered temperature trends (including: average, minimum, maximum and diurnal temperature range for different locations). The studies about temperature have also considered trends in temperature with other parameters. For instance, Bloomer et al. (2010) found a positive correlation between ozone (O3) and temperature. Frankes et al. (1978) have presented a study about temperature trends in relation to circulation. They found that in relation to pressure-trend, temperature trends were related to circulation changes.

Temperature trends may not follow a simple linear trend (fig. 2.1). There are mechanisms in the climate system both natural and anthropogenic which impact on the variations in climate trends. IPCC (4th assessment) report has for instance explained that; with regards to the climate response to volcanic aerosol, these forcings can directly affect the Earth’s radiative balance and thus alter surface temperature. According to IPCC (4th Assessment) Report, the overall effect of anthropogenic land cover change on global temperature will depend largely on the relative importance of increased surface albedo in winter and spring (exerting a cooling) and reduced evaporation in summer and in the tropics (exerting a warming). On the other hand, Wang et al. (2011) have reported that; both the surface solar radiation and atmospheric visibility in China exhibit a decreasing trend from 1961 to 1990. They have also observed that the decadal surface global radiation variation does not match continued surface temperature increases in most regions of China.

DISSERTATION SUBMITTED TO NANJING UNIVERSITY OF INFORMATION SCIENCE & TECHNOLOGY

Source: IPCC (4th assessment) Report

Figure: 2.1 Global temperature trends.

IPCC (4th assessment) report has revealed that; many scholars have carried out extensive research about temperature trends. These scholars5 include: Köppen, Callendar, Willett, Mitchell, Budyko, Jones et al., Hansen and Lebedeff, Brohan et al. The figure (Fig. 2.1) presents the results that were obtained by the scholars regarding global temperature trends.

Other studies conducted by other scholars have found different regional-level variations in the trends of both maximum and minimum temperatures, resulting in varying trends in diurnal temperature range (DTR) (Roy et al., 2009; Akinremi et al., 1999). The National Research Council (NRC) of US (2000) has indicated that daily minimum temperature have been rising more rapidly than daily maximum temperature which translated to decrease in DTR. The observations of NRC

5 The scholars identified above have used different methods and data to study temperature trends over wide spread region and details of their methods have been given in IPCC (4th assessment) report.

CHAPTER TWO LITERATURE REVIEW where also confirmed by Akinremi et al., (1999) in their study of precipitation trends in Canadian Prairies. Wang et al. (2011), noted that: changes in clouds, aerosols, urbanization and solar radiation affect could affect DTR

2.3 Precipitation

Hanif (2005) has defined variability of precipitation as ‘deviation of mean annual precipitation of a certain area from its amount calculated for a long period of time’. It was further noted that average amount of precipitation in a certain region differs from year to year and therefore from its long-term mean. It is this difference that is called variability. Where precipitation has appeared to fluctuate (vary), Laity (2008)) has blamed it on the fluctuation in temperature of seas which are the major source of moisture in the atmosphere. On the other hand Lau and Min (2006) have explained that solar radiation, air temperature, and winds are among the key parameters for determining precipitation. This therefore means, if temperature changes, there are chances of precipitation also changing.

The amount of precipitation received on different areas of the globe is not uniform (fig. 2.2). To appreciate the importance of precipitation, the amount and timings of precipitation cannot be ignored. For example; Akinremi et al., (1999) explained the importance of amount and timing of precipitation to grain production in the Canadian Prairies. There is also considerable variation in both temporal and spatial occurrence of precipitation (Feuerbacher and Stoewer, 2006; Michaelides et al., 2009; Wang and Cho, 1997). For instance; due to intense solar heating of the tropics and circulation properties, the equatorial regions normally receive high rainfall amounts (Michaelides et al., 2009) compared to the subtropical, mid-latitudes and polar region. Michaelides et al., 2009 also explain the importance of land-ocean distribution in which case, continental areas are usually drier because of their distance from moisture sources

DISSERTATION SUBMITTED TO NANJING UNIVERSITY OF INFORMATION SCIENCE & TECHNOLOGY

Source: http://www.eoearth.org/article/Global_distribution_of_precipitation

Figure 2.2: Global precipitation climatology.

In regard to climate variability, Baas and Selvaraju (2007) have further explained that some years may for instance have below average rainfall, yet some have average or even above average rainfall – this is does not mean climate change. Akinremi et al. (1999) have also argued that there is an inherent spatial and temporal variability of precipitation events and such variability should not be considered as climate change. In this regard, Wang and Cho (1997) explained the importance of having knowledge of precipitation variability such as trends and periodicity in regard to understanding the earth’s climate system. There is seasonality in the occurrence of extreme precipitation and floods (Stefan et al., 2002). It can therefore be accepted that, if we can predict extreme precipitation, considering the precipitation trends, we can be in position to predict occurrence of floods as well.

As for China, Gemmer et al. (2011) have reported that there have been changes in annual precipitation for the last century. They have also noted that in the last

CHAPTER TWO LITERATURE REVIEW decade there was an increase in national average precipitation with increasing summer precipitation while decreasing for autumn.

2.4 Diurnal temperature range and precipitation

Akinremi et al. (1999) have explained the indirect relationship between DTR and precipitation. It has been evident from previous studies that DTR has been reducing whereby minimum temperatures over most parts have been increasing faster than maximum temperatures (Akinremi et al. 1999; Bonsal et al. 2001; NRC of US, 2000; Wang et al. 2011). Other studies have indicated that what caused this trend in DTR were changes in clouds, aerosols, urbanization and solar radiation (Akinremi et al. 1999; Wang et al., 2011). Akinremi et al. (1999) have explained that increasing cloud cover may be associated with precipitation events since cloud cover is a precursor of precipitation.

2.5 Remaining problem: Comment on literature review

Having reviewed the available literature, this study considers specific location in the northern part of China, including the north western and north eastern areas as the study regions. It is true that there are available trends in terms of both temperature and precipitation either globally or regionally that have been discovered but this study considers the above regions with an underlying idea that: ‘how are the climate responses for urban and desert areas related?’ which response is greater or which one is smaller? Further study of the variability and trend of climate extremes has been studied. The climate extremes studied take consideration of: (1) high temperature days which are the days for which maximum temperature is greater than the threshold, 35oC, (2) high temperature extremes which are the days where the high temperature is greater than 95th percentile, (3) frost days were these are the days with their minimum temperature less than 0oC, (4) low temperature extremes which are the days with low temperature less than 5th percentile and (5) precipitation extremes which are the days with rainfall greater than 99th percentile.

DISSERTATION SUBMITTED TO NANJING UNIVERSITY OF INFORMATION SCIENCE & TECHNOLOGY CHAPTER THREE

3.0 DATA AND METHODOLOGY

This chapter presents the data used in the study and the methods used to analyze the data to arrive at the results which have been obtained. Section 3.1 has been dedicated to present the type of data used and source thereof. Section 3.2 has presented the relevant methods used to analyze data for the study.

3.1 Data used in the study

3.1.1 Temperature data

Daily temperature data for 20stations was obtained from School of Atmospheric Science of NUIST. This data is regularly updated and quality controlled to take care of the on-going research in the university. Of the 20stations, 10stations were for urban cities and 10stations for desert areas all located in the northern part of China. Details of these areas and stations have been presented in chapter one (section 1.4).

The temperature data was organized into annual and seasonal (spring and summer) classification. This data was further classified as for maximum temperature and minimum temperature. The difference between daily maximum temperature and daily minimum temperature gave the diurnal temperature range (DTR), which was also organized as annual and seasonal for analysis.

DTR  Tmax - T min (3.1)

Where: Tmax represent daily maximum temperature, and Tmin represents daily minimum temperature.

The temperature data was further analysed and temperature extremes extracted from the data. These temperature extremes included: (1) the number of days with maximum temperatures over 35.0oC (High Temperature Days), (2) the number of days with minimum temperatures below 0oC (Frost Days), (3) the number of days with maximum temperatures over the 95th percentiles of daily maximum

CHAPTER THREE DATA AND METHODOLOGY temperatures (High Temperature Extremes), and (4) the number of days with minimum temperatures below the 5th percentiles of daily minimum temperatures (Low Temperature Extremes).

3.1.2 Precipitation data

The precipitation data used, was obtained on daily basis. From the daily data, total annual precipitation (TOT-prcp) and rainy days (R.Days) were obtained. The precipitation data was also organised in terms of seasons again as amount for the season and the rain days for that season. The data for precipitation was further analysed in terms of the number of days with daily precipitation over the 99th percentiles of daily precipitation (Precipitation Extremes).

3.2 Methods used in the study

The study period was from 1981 to 2010 with efforts directed to study season (spring and summer) as well as annually. According to Lau and Mink (2006), trend analyses of rainfall (precipitation) deal with variations with time. The methods which were used in this study are: (1) the relative deficit (or surplus, subsection: 3.2.1), (2) coefficient of variation (sub section: 3.2.2), (3) the Mann- Kendall trend test (sub section: 3.2.3) and (4) the Beard formula for climate extremes (sub section: 3.2.4). Details of these methods have been presented in respective sub sections that follow.

3.2.1 The relative deficit or surplus (relative anomalies)

In order to study the trends, bearing in mind that different places receive different amounts of precipitation and have different temperatures due to different climatology, relative deficit and or surplus was calculated using the equation below:

R  R Relative deficit (or surplus)  i (3.2) R

Where: Ri and R are the data (temperature or precipitation) and their respective average over the study period. The average, x of a given set of data; x1, x2, ..., xN is computed from:

DISSERTATION SUBMITTED TO NANJING UNIVERSITY OF INFORMATION SCIENCE & TECHNOLOGY N x12 x ,...,  xN 1 x  or xx  i (3.3) N N i1

A graph has been plotted for the relative deficit (or surplus) of a given quantity (temperature or precipitation) against the study period. Such graph enables to give a snapshot of the deficient and surplus periods over the study regions in regard to the element being studied.

3.2.2 Standard deviation and coefficient of variation

Standard deviation6 is one of the measures that can be used to calculate spread in a set of data. When a lot of data is clustered around the mean, so that the distribution is approximate normal distribution, the standard deviation is small7. Standard deviation was used to obtain the spread of temperature and precipitation sets of data. The standard deviation of a set of data: X1, X2, ..., XN is calculated using:

N 1 2 Standard deviation,   (i ) (3.4) N i1

The calculated standard deviation was then used to obtain coefficient of variation. Coefficient of variation was more preferred to standard deviation because, for instance, two stations could have say same standard deviation regarding precipitation yet one has higher annual amount of rainfall compared to the other. To solve this problem8, coefficient of variation is used. Coefficient of variation is therefore a relative measure. The major limitation of coefficient of variation, however, is when the mean is zero or approaches zero.

The coefficient of variation9 as a percentage is given by the ratio of standard deviation to mean of the data set. Coefficient of variation, Vδ, is therefore obtained mathematically using:

 V 100% (3.5)  x

6 http://users.rcn.com/jkimball.ma.ultranet/BiologyPages/S/StandardError.html 7 http://www.robertniles.com/stats/stdev.shtml 8 http://www.vias.org/tmdatanaleng/cc_coeff_variation.html 9 http://www.westgard.com/lesson34.htm

CHAPTER THREE DATA AND METHODOLOGY

3.2.3 The Mann-Kendall trend test

The Mann–Kendall (M–K) trend test method was used to analyze trends the trends of rainfall amount, rain days, maximum, minimum temperature and DTR. The reasons for using this trend test have been given by Zhang et. al. (2011) which include the following reasons: (1) It is a rank-based nonparametric M–K test which can test trends without requiring normality or linearity. (2) This method, like other non-parametric trend detection methods, is less sensitive to outliers than are parametric statistics. (3) This method has been recommended by the World Meteorological Organization.

Jagannathan et al. (1974) have identified the Mann-Kendall rank statistic as a powerful test for trends that are linear or non-linear. This trend test has been used to test for trends in this study. Wilks (2011) defines Mann-Kendall trend test using the formula below:

nn1 Ssgn( xji x ) (3.6) i11 j  i 

Where:

1, (xxji  )  0  sgn(xj x i )  0, ( x j  x i )  0 (3.7)  1, (xxji  )  0

If the values are not repeated, then the variance, δ2(S) of the sampling distribution is given by:

n(n 1)(2n  5)  2 (S)  (3.8) 18

The test statistic is then given by the standard Gaussian value, z defined as:

DISSERTATION SUBMITTED TO NANJING UNIVERSITY OF INFORMATION SCIENCE & TECHNOLOGY S 1 S  0  (S)    z  0 S  0  (3.9)  S 1  , S  0  (S)

3.2.4 The Beard formula for climate extremes

The formula introduced by Beard was used to obtain the climate extremes. This formula is discussed by Folland et. al. (2002) in comparison with other formulae for studying extremes and is recommended for the study of climate extremes (Bonsal et. al. 2001; Folland et. al. 2002). According to the Beard formula, the probability, p that a random value is less than or equal to the rank of that value,

Xm is given by:

m  0.31 P  (3.10) N  0.38

Where m is the position of the value and N is the number of values in the data set. The values (say temperature) in the data set for example: summer season which has 92days, is first arranged in ascending order; X1, X2, . . . ,X91, X92. The temperature representing the 95th percentile is linearly interpolated between the 88th ranked value (giving: P = 94.9%) and 89th ranked value (P = 96.0%). The 95th percentile is therefore interpolated from the above results.

In this study, the climate extremes studied are: high temperature days, frost days, high temperature extremes, low temperature extremes, and precipitation extremes. The high temperature days are defined as the number of days with maximum temperatures over 35.0oC., frost days are the number of days with minimum temperatures below 0oC., high temperature extreme days are the number of days with maximum temperatures over the 95th percentiles of daily maximum temperatures, low temperature extreme days are the number of days with minimum temperatures below the 5th percentiles of daily minimum temperatures.

CHAPTER THREE DATA AND METHODOLOGY

For precipitation extremes, these are days having daily precipitation over the 99th percentiles of daily precipitation.

DISSERTATION SUBMITTED TO NANJING UNIVERSITY OF INFORMATION SCIENCE & TECHNOLOGY CHAPTER FOUR

4.0 RESULTS AND DISCUSSION

In this chapter, results have been presented and discussed. They have been presented in accordance with the stated objectives. The first section (4.1), deals with the temperature response the second section (4.2) deals with the precipitation response.

4.1 Temperature response

Under this section, the trends of temperature over the study period are presented. They have been studied as maximum, minimum temperature and diurnal temperature range (DTR). The temperature trends for desert areas (sub section 4.1.1) and urban areas (sub section 4.1.2) have been presented and discussed.

4.1.1 Over desert areas

This sub section presents the results for temperature trends for selected desert areas in China considering three timescales, namely: annual trends (sub section 4.1.1.1), spring trends (sub section 4.1.1.2) and summer trends (sub section 4.1.1.3).

4.1.1.1 Annual temperature trends

Table 4.1 shows the trend of annual temperature for selected desert areas under the study. The results have been obtained using Mann-Kendall statistic at 99% confidence level.

CHAPTER FOUR RESULTS AND DISCUSSION

Table 4.1: M-K trend (z) results for annual temperature - Desert

Tmax Tmin DTR Colours Alza Zuoqi 0.370 0.527 -0.457 Increasing trend Bayan MOD 0.448 0.037 0.269 Decreasing trend Da-Qaidam 0.591 0.497 -0.067 Guaizihu 0.301 0.467 -0.223 Hami 0.444 0.144 0.301 Hoboksar 0.264 0.385 -0.269 Linhe 0.315 0.480 -0.545 Qiemo 0.522 0.406 0.278 Ruoqiang 0.545 0.531 0.177 Turpan 0.508 0.596 -0.375 Avg 0.431 0.407 -0.091

Results show that both maximum and minimum temperatures are increasing on annual scale. Some areas have increasing DTR while other areas have decreasing DTR. When taken on average, DTR is decreasing (M-K, z = -0.091). Generally both maximum temperature (M-K, z = 0.431) and minimum temperature (M-K, z = 0.407) are increasing

The figures (Fig. 4.1, Fig. 4.2 and Fig. 4.3) show that; annual DTR for the selected desert regions is almost constant but highly variable. Annual DTR peaked during the last decade of the 20th century and appearing to follow a mild decreasing trend in this decade (2001 – 2010). It is this variability that could explain the weak decreasing trend given obtained in table 4.1 above. Further analysis of the annual rate of change of DTR for individual stations revealed that the rate of decrease was in the range of: -0.35oC/decade to -0.04 oC/decade.

DISSERTATION SUBMITTED TO NANJING UNIVERSITY OF INFORMATION SCIENCE & TECHNOLOGY

Fig. 4.1: Annual DTR relative anomalies for selected desert areas.

Fig. 4.2: Annual maximum temperature relative anomalies for selected desert areas.

Maximum temperature appears to also follow a sharp increasing trend over the study period. Further analysis indicates that the annual rate of increase of maximum temperature for individual desert stations under study was in the range

CHAPTER FOUR RESULTS AND DISCUSSION of: 0.36 oC/decade to 0.70 oC/decade. When aggregated and averaged, the annual rate of increase of temperature for the desert stations was: 0.51 oC/decade.

The trend for minimum temperature is largely variable but also shows a mild increasing trend. However, further analysis using the individual trends of stations indicated that; the rate of increase of minimum temperature per year was in the range of: 0.06 oC/decade to 0.8 oC/decade. When aggregated and averaged over the entire stations, minimum temperature was increasing on average of 0.52oC/decade, a rate slightly greater than the rate of increase for maximum temperature. It can therefore be argued that minimum temperatures largely increase faster than maximum temperature, although the difference on annual scale is small which could be accounting for the small rate of decrease of DTR obtained above.

Fig. 4.3: Annual minimum temperature relative anomalies for selected desert areas.

4.1.1.2 Spring temperature trends

Table 4.2 shows the trend of spring temperature for selected desert areas under the study. The results have been obtained using Mann-Kendall statistic at 99% confidence level.

DISSERTATION SUBMITTED TO NANJING UNIVERSITY OF INFORMATION SCIENCE & TECHNOLOGY Temperature results for spring season indicate that; again both maximum and minimum temperatures are increasing. Some regions have their DTR increasing while others have it decreasing. When taken on average DTR for spring is decreasing (M-K, z = -0.050).

The results shown in figures (Fig. 4.4, Fig. 4.5 and Fig. 4.6) show that; spring DTR for the selected desert regions appears to increase from about the last 20years. Maximum temperature appears to also follow an increasing trend from the same period. The trend for minimum temperature is also following an increasing trend, although at a smaller rate than that of maximum temperature and still over the same period. This scenario could in part explain the mild increase of DTR.

Table 4.2: M-K trend (z) results for spring temperature - Desert

Tmax Tmin DTR Colours Alza Zuoqi 0.274 0.416 -0.218 Increasing trend Bayan MOD 0.329 0.195 0.149 Decreasing trend Da-Qaidam 0.467 0.301 0.122 Guaizihu 0.191 0.375 -0.260 Hami 0.255 0.269 0.113 Hoboksar 0.149 0.324 -0.209 Linhe 0.195 0.499 -0.370 Qiemo 0.411 0.425 0.297 Ruoqiang 0.343 0.508 0.090 Turpan 0.177 0.425 -0.209 Avg 0.279 0.374 -0.050

CHAPTER FOUR RESULTS AND DISCUSSION

Fig. 4.4: Spring DTR relative anomalies for selected desert areas.

Further analysis of DTR of individual stations showed that; the rate of decrease of DTR was in the range of -0.547oC/decade to -0.18oC/decade except for those stations that presented an increasing DTR trend. When aggregated and averaged over all the stations and study period, the rate of decrease of DTR was: - 0.023oC/decade (negative sign is maintained to emphasise the decrease and differentiate it from increasing trend, where increase is shown with a positive sign).

For maximum temperature trend, further analysis indicated that: different stations had different rates but all of them indicated increasing trend of maximum temperature. The rates of increase for maximum temperature for individual stations were in the range of: 0.3 oC/decade to 0.8 oC/decade. When aggregated and averaged over all the stations, the averaged trend was 0.54 oC/decade.

DISSERTATION SUBMITTED TO NANJING UNIVERSITY OF INFORMATION SCIENCE & TECHNOLOGY

Fig. 4.5: Spring maximum temperature relative anomalies for selected desert areas.

Fig. 4.6: Spring minimum temperature relative anomalies for selected desert areas.

The trend of spring minimum temperature (Fig. 4.6) appears to increase as well. Further analysis indicated that: individual stations had their spring minimum temperature increase at rates in the range of: 0.23oC/decade to 0.85 oC/decade.

CHAPTER FOUR RESULTS AND DISCUSSION

When aggregated and averaged, over the entire stations under study, the trend was 0.56 oC/decade.

4.1.1.3 Summer temperature trends

Table 4.3 shows the trend of summer temperature for selected desert areas under the study. The results have been obtained using Mann-Kendall statistic at 99% confidence level.

Results for summer still show that; both maximum and minimum temperatures are increasing. Half of the stations have their DTR increasing while the other half has DTR decreasing. When taken on average, the trend for DTR is negligible (M- K, z = -2.8x10-18).

Table 4.3: M-K trend (z) results for summer temperature - Desert

Tmax Tmin DTR Colours Alza Zuoqi 0.287 0.444 -0.301 Increasing trend Bayan MOD 0.343 0.292 0.278 Decreasing trend Da-Qaidam 0.439 0.513 -0.232 Guaizihu 0.434 0.410 0.062 Hami 0.485 0.195 0.315 Hoboksar 0.324 0.434 -0.090 Linhe 0.278 0.526 -0.315 Qiemo 0.494 0.343 0.159 Ruoqiang 0.522 0.462 -0.039 Turpan 0.499 0.333 0.163 -2.8x10- Avg 0.411 0.395 18

The results as shown in figures (Fig. 4.7, Fig. 4.8 and Fig. 4.9) show that summer DTR for the selected desert regions appears to decrease mildly. Further investigation, using regression analysis and considering individual stations, only

DISSERTATION SUBMITTED TO NANJING UNIVERSITY OF INFORMATION SCIENCE & TECHNOLOGY three stations had a decreasing trend. The other stations had a small increasing trend, ranging from: 0.026oC/decade to 0.382 oC/decade. The averaged trend for the entire stations was: 0.068 oC/decade.

Fig. 4.7: Summer DTR relative anomalies for selected desert areas.

The summer maximum temperature of the desert stations under study appears to follow a clear increasing trend (Fig. 4.8). Using regression analysis to analyse individual stations, all the stations had increasing trend of summer maximum temperature. The trend was in the range of: 0.369oC/decade to 0.723 oC/decade. The averaged rate over the entire desert stations under study was 0.55 oC/decade.

CHAPTER FOUR RESULTS AND DISCUSSION

Fig. 4.8: Summer maximum temperature relative anomalies for selected desert areas.

The trend for summer minimum temperature for the desert regions under study is also following a sharp increasing trend, especially for the last 15years. This trend is sharper than that of maximum temperature. Further analysis using regression indicated that all stations had their summer minimum temperature increasing at a rate in the range of: 0.28oC/decade to 0.94 oC/decade. On average, the averaged rate for the entire stations under study was: 0.536 oC/decade.

DISSERTATION SUBMITTED TO NANJING UNIVERSITY OF INFORMATION SCIENCE & TECHNOLOGY

Fig. 4.9: Summer minimum temperature relative anomalies for selected desert areas.

4.1.2 Over selected cities

This sub section presents the results for temperature trends for selected urban areas in China considering three timescales, namely: annual trends (sub section 4.1.1.1), spring trends (sub section 4.1.1.2) and summer trends (sub section 4.1.1.3).

4.1.2.1 Annual temperature trends

Table 4.4 shows the trend of annual temperature for selected urban areas (cities) under the study. The results have been obtained using Mann-Kendall statistic at 99% confidence level.

Results generally show that both maximum and minimum temperatures are increasing on annual scale with exception of just one station, Shenyang whose minimum temperature was following a decreasing trend. Further still, the results generally show that DTR is following a decreasing tend (M-K, z = -0.293) with the exception of Shenyang and Tianjin which had an increasing trend of DTR.

The results in figure (Fig. 4.10) show that; annual DTR for selected urban cities appears to decrease sharply. Further analysis of DTR, using regression analysis

CHAPTER FOUR RESULTS AND DISCUSSION and analysing for each station also revealed that only Shenyang and Tianjin had increasing trend of DTR. All the other urban areas (cities) stations under study and over the study period had a decreasing annual trend of DTR in the range of; - 0.659oC/decade to -0.146oC/decade. When aggregated and averaged, the averaged trend was also a decreasing trend of -0.238oC/decade.

Table 4.4: M-K trend (z) results for annual temperature - Cities

Tmax Tmin DTR Colours Beijing 0.260 0.545 -0.384 Increasing trend Changchun 0.269 0.416 -0.320 Decreasing trend Dalian 0.228 0.370 -0.315 Datong 0.402 0.554 -0.292 Harbin 0.260 0.582 -0.683 Hohhot 0.384 0.577 -0.503 Shenyang 0.140 -0.145 0.214 Taiyuan 0.407 0.697 -0.393 Tangshan 0.343 0.485 -0.398 Tianjin 0.214 0.016 0.145 Avg 0.291 0.410 -0.293

The maximum temperature of the areas under study and over the study period appears to follow a mild increasing trend (Fig. 4.11). Further analysis using regression indicates that all the stations have their maximum temperature increasing over the study period. The rate of increase is in the range of: 0.11oC/decade to 0.501 oC/decade. The averaged rate of increase of maximum temperature over the entire stations under study and over the study period was: 0.325 oC/decade.

DISSERTATION SUBMITTED TO NANJING UNIVERSITY OF INFORMATION SCIENCE & TECHNOLOGY

Fig. 4.10: Annual DTR relative anomalies for selected cities.

Fig. 4.11: Annual maximum temperature relative anomalies for selected cities.

The trend for minimum temperature is almost stable about the average (Fig. 4.12). This figure does not offer a good insight. So, using regression analysis and analysing the annual minimum temperature trends for individual stations, only

CHAPTER FOUR RESULTS AND DISCUSSION

Shenyang presented a decreasing trend. The other stations all presented an increasing trend of annual minimum temperature in the range of: 0.022oC/decade to 1.015 oC/decade. The averaged trend is also following an increasing trend with a rate of 0.566 oC/decade. It is therefore obvious that the rate of increase of annual minimum temperatures over the study stations and over the study period is greater than that of maximum temperature.

Fig. 4.12: Annual minimum temperature anomalies for selected urban areas.

4.1.2.2 Spring temperature trends

Table 4.5 shows the trend of spring temperature for selected urban city areas under the study. The results have been obtained using Mann-Kendall statistic at 99% confidence level.

The results still show that both maximum and minimum temperatures are increasing in spring season with exception of, Harbin whose maximum temperature is following a decreasing trend and still Shenyang’s minimum temperature was following a decreasing trend. Results further generally show that, DTR is following a decreasing tend (M-K, z = -0.209) with the exception of Shenyang which had an increasing trend of DTR.

DISSERTATION SUBMITTED TO NANJING UNIVERSITY OF INFORMATION SCIENCE & TECHNOLOGY

Table 4.5: M-K trend (z) results for spring temperature - Cities

Tmax Tmin DTR Colours Beijing 0.154 0.338 -0.195 Increasing trend Changchun 0.113 0.264 -0.278 Decreasing trend Dalian 0.214 0.320 -0.175 Datong 0.186 0.377 -0.195 Harbin -0.002 0.324 -0.511 Hohhot 0.246 0.457 -0.297 Shenyang 0.117 -0.039 0.039 Taiyuan 0.315 0.476 -0.228 Tangshan 0.163 0.384 -0.338 Tianjin 0.195 0.149 0.090 Avg 0.170 0.305 -0.209

The results in figures (Fig. 4.13) show that; Spring DTR for selected urban cities appears to decrease sharply. Further analysis using regression and analysis individual stations revealed that only Shenyang and Tianjin had increasing trend of spring DTR. All the other stations had decreasing trend of DTR in the range of; -0.829oC/decade to -0.157 oC/decade. On aggregate, when the rates of the individual stations are averaged, the rate was still decreasing at -0.307 oC/decade.

The trend for spring maximum temperature (Table 4.5, Fig. 4.14) of the urban areas under study is also increasing. Results from Mann-Kendall statistic present the averaged value at z = 0.170. Further analysis using regression also show that all stations had an increasing summer trend for the maximum temperature in the range of 0.052 oC/decade to 0.533 oC/decade. On aggregate, the averaged rate over the entire study period and the study stations, the rate was 0.252 oC/decade.

CHAPTER FOUR RESULTS AND DISCUSSION

Fig. 4.13: Spring DTR relative anomalies for selected cities.

Fig. 4.14: Spring maximum temperature relative anomalies for selected cities.

The trend for spring minimum temperature (Table 4.5, Fig. 4.15) is an increasing one over the study stations and the study period. When further analysis was done for individual stations, only Shenyang presented a decreasing trend of spring minimum temperature. The other stations presented an increasing trend of spring

DISSERTATION SUBMITTED TO NANJING UNIVERSITY OF INFORMATION SCIENCE & TECHNOLOGY minimum temperature in the range of 0.137 oC/decade to 0.928 oC/decade. On aggregate, the urban stations under study presented an increasing trend of 0.56 oC/decade. It is also clear that the rate at which spring minimum temperature (0.56 oC/decade ) is increasing appears to be slightly greater than the rate of increase for spring maximum temperature (0.252 oC/decade) which in part could explain the decreasing trend of spring DTR (-0.307 oC/decade).

Fig. 4.15: Spring minimum temperature relative anomalies for selected cities.

4.1.2.3 Summer temperature trends

Table 4.6 shows the trend of summer temperature for selected urban city areas under the study. The results have been obtained using Mann-Kendall statistic at 99% confidence level.

Results show that both maximum and minimum temperatures are following an increasing trend over the study period for the summer season with exception of Shenyang which has a decreasing minimum temperature over the same study period. The summer DTR is following a decreasing trend with the exception of Shenyang and Tianjin which have increasing DTR over the summer season and study period. Generally when taken on average, maximum temperature is

CHAPTER FOUR RESULTS AND DISCUSSION increasing (M-K, z = 0.238), minimum temperature is also increasing (M-K, z = 0.367) while DTR is decreasing (M-K, z = -0.109).

Considering figure (Fig. 4.16): Summer DTR for selected urban cities appears to follow a decreasing trend. Further analysis using regression showed that all the stations with exception of Shenyang and Tianjin had a decreasing trend of summer DTR in the range of: -0.419oC/decade to -0.023 oC/decade. The averaged trend for the entire stations in summer had DTR decreasing at a rate: -0.14 oC/decade.

Table 4.6: M-K trend (z) results for summer temperature - Cities Max.Temp Min.Temp DTR Colours Beijing 0.228 0.494 -0.255 Increasing trend Changchun 0.287 0.352 -0.136 Decreasing trend Dalian 0.085 0.186 -0.145 Datong 0.352 0.526 -0.011 Harbin 0.267 0.494 -0.287 Hohhot 0.324 0.536 -0.067 Shenyang 0.071 -0.057 0.182 Taiyuan 0.379 0.568 -0.140 Tangshan 0.237 0.384 -0.329 Tianjin 0.154 0.186 0.094 Avg 0.238 0.367 -0.109

DISSERTATION SUBMITTED TO NANJING UNIVERSITY OF INFORMATION SCIENCE & TECHNOLOGY

Fig. 4.16: Summer DTR relative anomalies for selected cities.

The summer maximum temperatures (Fig. 4.17) appear to follow an increasing trend especially in the last 20years (1991 to 2010). Further analysis for individual stations using regression analysis revealed that all the urban stations under study and over the study period had an increasing trend of summer maximum temperatures in the range of: 0.107oC/decade to 0.684oC/decade. When aggregated and averaged, the stations presented an increasing trend at a rate of 0.389 oC/decade.

Just like the summer maximum temperature, summer minimum temperature also followed an increasing trend (Fig. 4.18) especially from around 1993. When individual stations were further analysed using regression, all stations had increasing trend of summer minimum temperature with the exception of Shenyang. The stations had an increasing trend in the range of 0.144oC/decade to 0.864 oC/decade. When the stations are aggregated, the average rate of increase of summer minimum temperature was: 0.529 oC/decade. It is also clear that the rate at which minimum temperature is increasing appears to be greater than the rate of increase for maximum temperature again which in part could explain the decreasing trend of DTR.

CHAPTER FOUR RESULTS AND DISCUSSION

Fig. 4.17: Summer maximum temperature relative anomalies for selected cities.

Fig. 4.18: Summer minimum temperature relative anomalies for selected cities.

4.2 Precipitation response

Under this section, the trends of precipitation over the study period are presented. These trends have been studied as precipitation amount and rain days for desert

DISSERTATION SUBMITTED TO NANJING UNIVERSITY OF INFORMATION SCIENCE & TECHNOLOGY areas (sub section 4.2.1) and selected urban areas (sub section 4.2.2). The time scale considered for the precipitation trends has also been discussed in terms of annual, spring and summer respectively.

4.2.1 Over desert areas

This sub section presents the results for precipitation trends for selected desert areas in China considering three timescales, namely: annual trends (sub section 4.2.1.1), spring trends (sub section 4.2.1.2) and summer trends (sub section 4.2.1.3).

4.2.1.1 Annual precipitation trends

Table 4.7 shows the trend of annual precipitation for desert areas under the study. The results have been obtained using Mann-Kendall statistic at 99% confidence level.

Table 4.7: M-K trend (z) results for annual precipitation - Deserts TOT-prcp R.Days Colour Alza Zuoqi 0.080 0.002 Increasing trend Bayan MOD 0.407 0.223 Decreasing trend Da-Qaidam 0.149 0.131 Guaizihu 0.168 0.014 Hami 0.168 0.161 Hoboksar 0.113 0.137 Linhe 0.034 -0.110 Qiemo -0.030 0.097 Ruoqiang 0.145 -0.010 Turpan -0.011 0.262 Avg 0.122 0.091

From the results presented in the table 4.7 above, it is clear that; on annual scale, most of the stations have increasing trends of precipitation and rain days. Taking

CHAPTER FOUR RESULTS AND DISCUSSION the average of the stations, still it gives increasing trend of rainfall amount10 (M-K, z = 0.122) and increasing trend of rain days (M-K, z = 0.091). What is also clearer is that the rate at which rainfall is increasing as given by Mann-Kendall statistic is greater than that given for the rain days. Linhe and Ruoqiang have increasing rainfall yet decreasing rain days. This probably could imply that; the average rainfall intensity is increasing for these two stations. Indeed Zhai et. al. (2005) discussed this situation and concluded that average daily rain rate (intensity) may have increased. Qiemo and Turpan have decreasing precipitation and increasing rain days. This situation could probably indirectly indicate that average rainfall intensity is decreasing.

Although the rain days over the study period and on annual scale for the desert areas has been varying, results show that in the recent decade, the rain days have tended to be generally greater than the average. This is shown by the increasing trend shown by the figure (Fig. 4.19) below. The annual rainfall amount also is following an increasing trend over the last decade (Fig. 4.20) moreover the anomalies are greater compared to the deficits.

Fig. 4.19: Annual trend of rain days relative anomalies for desert regions.

10 Annual rainfall amount refer to the total precipitation received during that year which was obtained by summing the daily precipitation for that respective year.

DISSERTATION SUBMITTED TO NANJING UNIVERSITY OF INFORMATION SCIENCE & TECHNOLOGY

Fig. 4.20: Annual trend of rainfall amount relative anomalies for desert regions.

The results obtained above, which indicate increasing trend of annual precipitation for the selected desert areas are in agreement with Zhai et. al. (2005) who also established that annual precipitation amount has increased over western China.

4.2.1.2 Spring (MAM) precipitation trends

Table 4.8 shows the trend of precipitation for desert areas under the study during the spring season (March – May). The results have been obtained using Mann- Kendall statistic at 99% confidence level.

CHAPTER FOUR RESULTS AND DISCUSSION

Table 4.8: M-K trend (z) results for spring precipitation - Deserts TOT-prcp R.Days Colour Alza Zuoqi 0.138 0.130 Increasing trend Bayan MOD 0.299 0.282 Decreasing trend Da-Qaidam 0.044 -0.093 Guaizihu 0.164 -0.003 Hami 0.192 0.128 Hoboksar 0.285 0.021 Linhe 0.101 0.135 Qiemo 0.104 0.092 Ruoqiang 0.063 0.029 Turpan 0.207 0.292 Avg 0.160 0.101

From the results presented in the table 4.8 above, it is clear that: for spring season, all the stations have increasing trend of rainfall amount11 and only Da-Qiadam and Guaizhu have a decreasing trend of rain days. The other stations all have increasing trend of rain days. Taking the average of the stations, still it gives increasing trend of precipitation (M-K, z = 0.160) and increasing trend of rain days (M-K, z = 0.101). These results are in agreement with Zhai et. al. (2005), who noted that: spring precipitation has increased in many parts of western and southwest China. Although the rain days are decreasing in spring for the two stations named above, the precipitation is increasing. This could suggest that average rainfall intensity in the spring season is increasing.

11 Rainfall amount for spring season means total daily precipitation for the spring season. It was obtained by summing up the daily precipitation for that respective year and over the spring season (i.e. months of March – May)

DISSERTATION SUBMITTED TO NANJING UNIVERSITY OF INFORMATION SCIENCE & TECHNOLOGY

Fig. 4.21: Spring trend of rain days relative anomalies for desert regions.

Fig. 4.22: Spring trend of rainfall amount relative anomalies for desert regions.

Results obtained after investigating the anomalies (relative deficits/surplus) for both rainfall and rain days indicate that although the trend is appearing to increase, it is largely following a sinusoidal trend. The periodicities of these trends may

CHAPTER FOUR RESULTS AND DISCUSSION however need further investigation but they appear to be in the range of 10 – 13years as observed from the figures (fig. 4.21 and fig. 4.22).

4.2.1.3 Summer precipitation trends

Table 4.9 shows the trend of precipitation for desert areas under the study during the summer season (June – August). The results have been obtained using Mann- Kendall statistic at 99% confidence level.

The results presented in the table 4.9 below, have mixed situation with some stations having decreasing trend while others having increasing trend. Four stations have both decreasing rainfall amount12 and decreasing rain fall days. Three stations have increasing trend both for rainfall amount and rain days. When aggregated on average, results give a mild increasing precipitation trend (M-K, z = 0.014) and a small decreasing trend of rain days (M-K, z = -0.072). Such mixed results require further analysis using powerful numerical models to analyze them further and carryout projection into the future.

According to Mann-Kendall results (table 4.9), the rain days are following a mild decreasing trend (M-K, z = -0.072). When rain days anomalies are graphed (fig. 4.23), results show that although the rain days are decreasing as well, there is a variability where the period is decreasing as well. The variability for rainfall (fig. 4.24) is central about the zero and also the period appears to decrease.

12 Rainfall amount for the Summer season indicate the total daily precipitation summed up for the year during the summer season.

DISSERTATION SUBMITTED TO NANJING UNIVERSITY OF INFORMATION SCIENCE & TECHNOLOGY Table 4.9: M-K trend (z) results for summer precipitation - Deserts TOT-prcp R.Days Colour Alza Zuoqi -0.067 -0.239 Increasing trend Bayan MOD 0.228 -0.042 Decreasing trend Da-Qaidam 0.108 0.181 Guaizihu -0.074 -0.169 Hami 0.090 0.002 Hoboksar -0.011 0.009 Linhe -0.021 -0.254 Qiemo -0.196 -0.134 Ruoqiang 0.057 -0.221 Turpan 0.025 0.148 Avg 0.014 -0.072

Fig. 4.23: Summer trend of rain days relative anomalies for desert regions.

CHAPTER FOUR RESULTS AND DISCUSSION

Fig. 4.24: Summer trend of rainfall amount relative anomalies for desert regions.

4.2.2 Over selected cities

This sub section presents the results for precipitation trends for selected cities in China considering three timescales, namely: annual trends (sub section 4.2.2.1), spring trends (sub section 4.2.2.2) and summer trends (sub section 4.2.2.3).

4.2.2.1 Annual precipitation trends

Table 4.10 shows the trend of annual precipitation for selected urban areas under the study. The results have been obtained using Mann-Kendall statistic at 99% confidence level.

DISSERTATION SUBMITTED TO NANJING UNIVERSITY OF INFORMATION SCIENCE & TECHNOLOGY

Table 4.10: M-K trend (z) results for annual precipitation - Cities TOT-prcp R.Days Colour Beijing -0.145 -0.161 Increasing trend Decreasing Changchun -0.044 0.028 trend Dalian 0.154 0.019 Datong -0.048 -0.028 Harbin -0.228 -0.251 Hohhot -0.053 -0.061 Shenyang 0.057 -0.026 Taiyuan -0.021 0.059 Tangshan -0.149 -0.148 Tianjin -0.103 -0.071 Avg -0.058 -0.064

Fig. 4.25: Annual trend of rain days relative anomalies for selected cities

CHAPTER FOUR RESULTS AND DISCUSSION

The Results shown in table 4.10 above generally indicate that annual precipitation of the selected urban areas is decreasing as well as the rain days. Dalian has increasing precipitation and rain days. Shenyang has increasing precipitation but decreasing rain days. Changchun and Taiyuan both have decreasing precipitation but increasing rain days. When taken on aggregate, the precipitation is decreasing at a small rate (M-K, z = -0.058) as do the rain days (M-K, z = -0.64). These results are consistent with Zhai et. al. (2005) observation. They noted that annual trends of precipitation are not consistent in China as a whole. They noted that, it has significantly decreased in North China and the Sichuian basin.

Considering the relative deficits/surplus (relative anomalies) of rain days, for the last 15years, the rain days for urban cities, on annual scale, have been below normal (fig. 4.25) above. The rain fall amount has been largely variable (fig. 4.26) appearing to pick up and be greater than average as from about 2007

Fig. 4.26: Annual trend of rainfall amount relative anomalies for selected cities.

DISSERTATION SUBMITTED TO NANJING UNIVERSITY OF INFORMATION SCIENCE & TECHNOLOGY

4.2.2.2 Spring (MAM) precipitation trends

Table 4.11 shows the trend of spring precipitation for selected urban areas under the study. The results have been obtained using Mann-Kendall statistic at 99% confidence level.

Results generally show that for most areas, the amount of precipitation is increasing as do the rain days. Still for Shenyang, the amount of precipitation is increasing while the rain days are decreasing. Again for Taiyuan, the amount of precipitation is decreasing while the rain days are increasing (table 4.5, below). Tianjin has maintained its trend with the amount of precipitation decreasing as noted on the annual scale earlier above, in the same way the rain days also decreasing. Taken on aggregate, the amount of precipitation (M-K, z = 0.095) and rain days are both increasing (M-K, z=0.041).

Considering the results for the spring season further, it is clear that the average rain days and average rainfall for the selected urban cities that, they have been varying, but started to following an increasing trend from around 2005 (fig. 4.27 and fig. 4.28).

Table 4.11: M-K trend (z) results for spring precipitation - Cities TOT-prcp R.Days Colour Beijing 0.228 0.145 Increasing trend Changchun 0.062 0.083 Decreasing trend Dalian 0.306 0.057 Datong 0.039 0.091 Harbin 0.209 0.050 Hohhot 0.071 0.005 Shenyang 0.103 -0.087 Taiyuan -0.108 0.080 Tangshan 0.044 0.033 Tianjin -0.007 -0.045 Avg 0.095 0.041

CHAPTER FOUR RESULTS AND DISCUSSION

Fig. 4.27: Spring trend of rain days relative anomalies for selected cities.

Fig. 4.28: Spring trend of rainfall amount relative anomalies for selected cities.

DISSERTATION SUBMITTED TO NANJING UNIVERSITY OF INFORMATION SCIENCE & TECHNOLOGY

4.2.2.3 Summer precipitation trends for selected urban cities

Table 4.12 shows the trend of summer precipitation for selected urban areas under the study. The results have been obtained using Mann-Kendall statistic at 99% confidence level.

Table 4.12: M-K trend (z) results for summer precipitation - Cities TOT-prcp R.Days Colour Beijing -0.264 -0.252 Increasing trend Changchun 0.002 -0.176 Decreasing trend Dalian 0.011 0.002 Datong -0.255 -0.112 Harbin -0.299 -0.182 Hohhot -0.191 -0.171 Shenyang 0.025 0.054 Taiyuan 0.080 -0.108 Tangshan -0.287 -0.257 Tianjin -0.154 -0.104 Avg -0.133 -0.131

The results generally show that; for most cities under study, both the amount of precipitation and rain days are decreasing (table 4.12). Dalian and Shenyang both have increasing trend of summer precipitation and rain days. While Changcun and Taiyuan both have increasing summer precipitation and decreasing number of summer rain days. Tianjin has maintained a decreasing trend of both rainfall and rain days throughout the seasons and the annual scale. When taken on aggregate, both summer precipitation (M-K, z = -0.133) and rain days (M-K, z = -0.131) are decreasing.

CHAPTER FOUR RESULTS AND DISCUSSION

Fig. 4.29: Summer trend of rain days relative anomalies for selected cities.

Fig. 4.30: Summer trend of rainfall amount relative anomalies for selected cities.

For the summer season, although the average rain days and average rainfall over the selected cities are highly variable, they are following a decreasing trend (fig. 4.29 and fig. 4.30). The trends are more evident in the last 13years, from around

DISSERTATION SUBMITTED TO NANJING UNIVERSITY OF INFORMATION SCIENCE & TECHNOLOGY 1997. These results for decreasing summer rain fall are in agreement with Meehl et. al. (2000) who noted that: there is a general drying of the midcontinental areas during summer. They attributed to increased temperature and evaporation along and argued that it could lead to drought.

4.3 Climate extremes

4.3.1 Over desert areas

4.3.1.1 High temperature days and high temperature extremes

Table 4.13 show the number of high temperature days and high temperature extreme days over desert areas under study. The number of high temperature days13, with exception of Da-Qaidam has been increasing for all the stations (M- K, z = 0.151 to 0.492) as well as the days experiencing high temperature extreme (M-K, z = 0.159 to 0.493). The highest temperatures for Da-Qaidam in summer were below 35oC and no wonder it also has the smallest rate of high temperature extremes. Correlation analysis between summer average maximum temperature over the study period and the high temperature extreme trend, however gives a moderate positive correlation coefficient of 0.424. The correlation coefficient between the high temperature days and high temperature extreme days is very weak at 0.228. Considering the summer maximum temperatures in general (sub section: 4.1.1.3), all the stations had increasing trend of summer maximum temperature.

13 The high temperature days are the number of days with maximum temperatures over 35.0oC

CHAPTER FOUR RESULTS AND DISCUSSION

Table 4.13: M-K trend (z) results for high temperature days and extremes – deserts High Temp. Days High Temp. Ext. Alza Zuoqi 0.250 0.493 Bayan MOD 0.338 0.440 Da-Qaidam N/A 0.159 Guaizihu 0.464 0.426 Hami 0.492 0.445 Hoboksar 0.151 0.308 Linhe 0.189 0.328 Qiemo 0.363 0.300 Ruoqiang 0.509 0.339 Turpan 0.386 0.353

4.3.1.2 Frost days and low temperature extremes

Table 4.14 show the number of frost days and days for low temperature extreme14 during the spring seasons for desert areas under study and over the study period. All the stations have a decreasing trend of frost days15 in the range of (M-K, z = - 0.414 to -0.128). The low temperature extreme is also generally decreasing (M-K, z = -0.446 to -0.010). The rates of decrease of frost days and low temperature extreme days have a strong positive correlation of 0.614. This decreasing trend means that few and fewer days have minimum temperature below zero degrees Celsius (0oC). This situation is further supported by the results which have been obtained in section (4.1.1.2) above where the trend of spring minimum temperature is increasing.

14 The low temperature extreme days are those days with their minimum temperature less than 5th percentile 15 The frost days are the number of days with minimum temperatures below 0oC.

DISSERTATION SUBMITTED TO NANJING UNIVERSITY OF INFORMATION SCIENCE & TECHNOLOGY Table 4.14: M-K trend (z) results for frost days and low temperature extremes –deserts Frost Days L. Temp. Ext. Alza Zuoqi -0.335 -0.209 Bayan MOD -0.128 -0.010 Da-Qaidam -0.355 -0.050 Guaizihu -0.375 -0.222 Hami -0.215 -0.107 Hoboksar -0.379 -0.094 Linhe -0.383 -0.446 Qiemo -0.155 -0.054 Ruoqiang -0.414 -0.242 Turpan -0.339 -0.125

4.3.1.3 Precipitation extremes

Table 4.15 shows the trend of number of days with precipitation extremes16 during the spring and summer seasons for the desert areas and over the study period. Just as in section 4.2.1.2, all stations presented an increasing trend of spring precipitation extreme (M-K, z = 0.003 to 0.292) with an average of 0.170. The incident of precipitation extremes in summer season are generally also increasing but at a smaller rate compared to the spring season and also three stations presented a decreasing trend of precipitation extremes. Taking the average for the summer season, the precipitation extremes are increasing at a rate of 0.002 (M-K, z = 0.002).

16 Precipitation extreme is defined as when the amount of rainfall is greater than 99% percentile of the climatological rainfall for the station or region.

CHAPTER FOUR RESULTS AND DISCUSSION

Table 4.15: M-K trend (z) results for precipitation extremes – deserts Spring Summer Alxa Zuoqi 0.251 -0.064 Bayan MOD 0.192 0.207 Da-Qaidam 0.003 0.069 Guaizihu 0.292 0.000 Hami 0.256 0.080 Hoboksar 0.232 0.048 Linhe 0.140 0.033 Qiemo 0.145 -0.319 Ruoqiang 0.080 0.045 Turpan 0.107 -0.079 Avg 0.170 0.002

4.3.2 Over selected cities

4.3.2.1 High temperature days and high temperature extremes

Table 4.16 shows the number of high temperature days and high temperature extremes for selected urban cities. Like for the desert areas, the number of high temperature days has been increasing for all the stations (M-K, z = 0.140 to 0.511). The number of days with high temperature extremes have also been increasing (M-K, z = 0.105 to 0.442). Correlation analysis between summer average maximum temperature over the study period and the high temperature extreme trend, also gives a moderate positive correlation coefficient, but slightly greater than that of deserts, of 0.488. The rates of increase of high temperature days and high temperature extreme days have a strong positive correlation of 0.683. This coefficient probably suggests that, general warming of summer is season is to be expected and that a warm summer season is to be characterised with also high temperature days. Considering the summer maximum temperatures of the selected cities in general (sub section: 4.1.2.3), all the stations had increasing trend of summer maximum temperature.

DISSERTATION SUBMITTED TO NANJING UNIVERSITY OF INFORMATION SCIENCE & TECHNOLOGY Table 4.16: M-K trend (z) results for high temperature days and extremes – cities

High Temp. Days High Temp. Ext. Beijing 0.283 0.343 Changchun 0.314 0.138 Dalian 0.179 0.105 Datong 0.505 0.321 Harbin 0.269 0.307 Hohhot 0.289 0.303 Shenyang 0.140 0.020 Taiyuan 0.511 0.442 Tangshan 0.242 0.307 Tianjin 0.252 0.327

4.3.2.2 Frost days and low temperature extremes

Table 4.14 show the number of frost days and low temperature extremes during the spring seasons for selected urban cities and over the study period. The trend of frost days for the selected cities is also decreasing. All the stations in the selected cities studied showed a decreasing trend in the range of (M-K, z = -0.362 to - 0.080) with exception of Shenyang (M-K, z = 0.021). The number of days having low temperature extremes is also decreasing (M-K, z = -0.442 to -0.148) with exception of Tianjin (M-K, z = 0.010). The rates of decrease of frost days and low temperature extreme days have a strong positive correlation of 0.730. Again this decreasing trend means that few and fewer days have minimum temperature below zero degrees Celsius (0oC). This situation is further supported by the results which have been obtained in section (4.1.2.2) above where the trend of spring minimum temperature is increasing.

What is also notable and clear is that the rate at which the frost days for desert areas is decreasing is greater than the rate for the selected urban areas.

CHAPTER FOUR RESULTS AND DISCUSSION

Table 4.17: M-K trend (z) results for frost days and low temperature extremes –cities Frost Days L. Temp. Ext. Beijing -0.125 -0.304 Changchun -0.133 -0.381 Dalian -0.231 -0.148 Datong -0.334 -0.311 Harbin -0.272 -0.442 Hohhot -0.324 -0.290 Shenyang 0.021 0.319 Taiyuan -0.362 -0.389 Tangshan -0.272 -0.184 Tianjin -0.080 0.010

4.3.2.3 Precipitation extremes

Table 4.18 shows the trend of number of days with precipitation extremes during the spring and summer seasons for the selected urban areas and over the study period. Again just as in section 4.2.2.2, all stations presented an increasing trend of spring precipitation extreme (M-K, z = 0.003 to 0.236) with an average of 0.030. Taiyuan here is the exception which has a decreasing trend of precipitation extremes (M-K, z = -0.287). The incident of precipitation extremes in summer season are generally decreasing with only three stations as exception (Changchun, Shenyang and Taiyuan). When aggregated and taken on average for the summer season, the precipitation extremes are decreasing at a rate of -0.118 (M-K, z = - 0.118).

Although the decreasing precipitation extremes in summer may appear to say bring the precipitation to acceptable levels, results obtained in section 4.2.2.3 showed that generally summer precipitation was decreasing. What this probably means is that coupled with the high temperatures of summer, there is a greater probability of drought and water shortage.

DISSERTATION SUBMITTED TO NANJING UNIVERSITY OF INFORMATION SCIENCE & TECHNOLOGY

Table 4.18: M-K trend (z) results for precipitation extremes – cities Spring Summer Beijing 0.003 -0.261 Changchun 0.051 0.087 Dalian 0.039 -0.202 Datong 0.061 -0.160 Harbin 0.236 -0.256 Hohhot 0.014 -0.190 Shenyang 0.079 0.189 Taiyuan -0.287 0.011 Tangshan 0.003 -0.290 Tianjin 0.102 -0.109 Avg 0.030 -0.118

CHAPTER FOUR RESULTS AND DISCUSSION

DISSERTATION SUBMITTED TO NANJING UNIVERSITY OF INFORMATION SCIENCE & TECHNOLOGY CHAPTER FIVE

5.0 CONCLUSION AND RECOMMENDATIONS

In this chapter, the results which were obtained in Chapter four have been summarised upon which conclusions and recommendation have been made. The first section 5.1 deals with temperature trends, the second section 5.2 deals with the precipitation trends and section 5.3 summarises climate extremes. The last section 5.4, deals with the summary and recommendations.

5.1 Temperature response

5.1.1 DTR trends

Table 5.1 summarises the results of DTR trend obtained by Man-Kendall statistics while table 5.2 summarises the results of DTR trend obtained using regression analysis. The results summarised in both tables were presented in Chapter 4 (section 4.1). Generally DTR is decreasing for both desert areas and urban cities. It is also evident that DTR for the urban cities is decreasing at a faster rate than that for the desert areas. The rate of decrease for DTR on annual scale was greater than for summer but smaller than for spring.

Average results for DTR using Table 5.1 M-K trend (z) test Annual Spring Summer Desert -0.091 -0.050 Not Sign17. Urban cities -0.293 -0.209 -0.109

Table 5.2 DTR – Regression Annual Spring Summer oC/decade oC/decade oC/decade Desert No Sign. -0.023 0.068 Urban cities -0.238 -0.307 -0.140

17 Not Sign. Means not significant as the magnitude of the value is too small for comparison and analysis.

CHAPTER FIVE CONCLUSION AND RECOMMENDATIONS

5.1.2 Maximum temperature trends

Table 5.3 summarises the results of maximum temperature trend obtained by Man-Kendall statistics while table 5.4 summarises the results of maximum temperature trend obtained using regression analysis. The results summarised in both tables were presented in Chapter 4 (section 4.1). Generally maximum temperature is increasing for both desert areas and urban cities. It is also evident that maximum temperature for the desert areas is increasing at a faster rate than that for the urban cities.

Average results for maximum Table 5.3: temperature using M-K trend (z) test Annual Spring Summer Desert 0.431 0.279 0.411 Urban cities 0.291 0.170 0.238

Table 5.4 Maximum temperature trend – regression Annual Spring Summer oC/decade oC/decade oC/decade Desert 0.510 0.540 0.550 Urban cities 0.325 0.252 0.389

5.1.3 Minimum temperature trends

Table 5.5 summarises the results of minimum temperature trend obtained by Man-Kendall statistics while table 5.6 summarises the results of minimum temperature trend obtained using regression analysis. The results summarised in both tables were presented in Chapter 4 (section 4.1). Generally minimum temperature is increasing for both desert areas and urban cities. It is also evident that minimum temperature for the desert areas is increasing at a faster rate than that for the urban cities for summer and spring seasons compared to annually.

DISSERTATION SUBMITTED TO NANJING UNIVERSITY OF INFORMATION SCIENCE & TECHNOLOGY

Average results for minimum Table 5.5 temperature using M-K trend (z) test Annual Spring Summer Desert 0.407 0.374 0.395 Urban cities 0.410 0.305 0.367

Table 5.6 Minimum temperature trend – regression Annual Spring Summer oC/decade oC/decade oC/decade Desert 0.520 0.560 0.536 Urban cities 0.566 0.560 0.529

5.2 Precipitation response

5.2.1 Trends of rainfall amounts

Table 5.7 summarises the results of rainfall amount obtained by Man-Kendall statistics and which were presented in Chapter 4 (section 4.2). Generally the rainfall amount is increasing for desert areas at a rate greater than the selected urban cities. The increase is more significant in the spring season than the annual scale and the summer seasons. For the urban cities, the rainfall amount is even decreasing in annual scale and summer season with a slight increase in the spring season which is still less than the rate for the desert areas.

Average results for rainfall Table 5.7 trend using M-K trend (z) test Annual Spring Summer Desert 0.120 0.160 0.014 Urban cities -0.058 0.095 -0.133

CHAPTER FIVE CONCLUSION AND RECOMMENDATIONS

5.2.2 Trends of rain days

Table 5.8 summarises the results of rain days obtained by Man-Kendall statistics and which were presented in Chapter 4 (section 4.2). Generally the rain days are increasing for desert areas at a rate greater than the selected urban cities. The increase is more significant in the spring season than the annual and the summer seasons. Further for the desert areas, the summer season is witnessing decreasing rain days. For the urban cities, the rain days are even decreasing in annual scale and summer season with a slight increase in the spring season which is still less than the rate for the desert areas.

Average results for rain Table 5.8 days trend using M-K trend (z) test Annual Spring Summer Desert 0.091 0.101 -0.072 Urban cities -0.064 0.041 -0.131

Zhai et. al. (2005) have argued that: the decrease in total precipitation in northern China (the selected urban cities, for this study) is due to the decrease in the number of rain days such that an increase in the average intensity was not able to compensate the loss of precipitation due to fewer rain days. In western China (which is predominantly the desert region as for this study), they further argue that: the increase in total precipitation over western China seems to be due to an increase in both precipitation frequency and average intensity.

Because of the difference in rates of precipitation, this study is in agreement with the results presented by Zhai et. al. (2005) that: annual precipitation does not show any significant trend for China as a whole; however, there are distinctive regional patterns.

DISSERTATION SUBMITTED TO NANJING UNIVERSITY OF INFORMATION SCIENCE & TECHNOLOGY 5.3 Climate extremes

In this subsection, the summaries of results regarding climate extremes have been presented. The first sub section (5.3.1) presents high temperature days and high temperature extremes, second subsection (5.3.2) presents frost days and low temperature extremes and the last subsection (5.3.3) presents precipitation extremes

5.3.1 High temperature days and high temperature extremes

Table 5.9 summarises the results obtained by Mann-Kendall regarding the trend of high temperature days and high temperature extremes. Generally both the desert areas and urban cities are exhibiting an increasing trend of high temperature days and high temperature extremes which supports results noted earlier in section 5.1.2. The rate of increase of high temperature extremes of desert areas is slightly higher than that of the urban cities. But generally, considering the summer maximum temperatures in general, all the stations show increasing trend of summer maximum temperature.

Table 5.9: Average results for high temperature days and high temperature extremes trend using M-K trend (z) test High Temp. High Temp. Days Ext. Desert areas 0.151 to 0.492 0.159 to 0.493 Urban cities 0141 to 0.511 0.105 to 0.442

5.3.2 Frost days and low temperature extremes

Table 5.10 summarises the results obtained by Mann-Kendall regarding the trend of frost days and low temperature extremes. Generally both the desert areas and urban cities are exhibiting a decreasing trend of frost days and low temperature extremes which supports results noted earlier in section 5.1.3.

CHAPTER FIVE CONCLUSION AND RECOMMENDATIONS

Table 5.10: Average results for frost days and low temperature extremes trend using M-K trend (z) test Frost Days L. Temp. Ext. Desert areas -0.414 to -0.128 -0.446 to -0.010 Urban cities -0.362 to -0.080 -0.442 to -0.148

The rate of decrease of frost days of desert areas is slightly higher than that of the urban cities. But generally, considering the spring minimum temperatures in general, all the stations show increasing trend of spring minimum temperature and that is why the frost days and low temperature extremes are decreasing.

5.3.3 Precipitation extremes

Table 5.11 summarises the results obtained by Mann-Kendall regarding the trend spring and summer precipitation extremes. It shows the average rate of precipitation extremes. Generally both the desert areas and urban cities are exhibiting an increasing trend of spring precipitation extremes. The desert areas have a slightly greater rate than urban cities. For summer season, the rate decreases and in fact the urban cities have a decreasing trend of precipitation extremes for summer while the desert have a small rate with some stations having a decreasing rate of precipitation extremes. This situation generally implies increased incidence of droughts in urban cities during the summer season and thus careful utilisation of water resources should be promoted.

Average results for Precipitation Table 5.11 extremes trend using M-K trend (z) test Spring Summer Desert areas 0.170 0.002 Urban cities 0.030 -0.118

DISSERTATION SUBMITTED TO NANJING UNIVERSITY OF INFORMATION SCIENCE & TECHNOLOGY 5.4 Summary and Recommendations

In summary, DTR trend is decreasing for both desert areas and urban cities. DTR for the urban cities is decreasing at a faster rate than that for the desert areas. The rate of decrease for DTR annually is greater than for both spring and summer seasons. The maximum temperature for the desert areas is increasing at a faster rate than that for the urban cities. For high temperature days and high temperature extremes, both the desert areas and urban cities are exhibiting an increasing trend. The rate of increase of high temperature extremes of desert areas is slightly higher than that of the urban cities. The minimum temperature is also increasing for both desert areas and urban cities. Minimum temperature for the desert areas is increasing at a faster rate than that for the urban cities for summer and spring seasons compared to annually. As for frost days and low temperature extremes, both the desert areas and urban cities are exhibiting a decreasing trend. The rate of decrease of frost days of desert areas is slightly higher than that of the urban cities.

For precipitation, the rainfall amount is increasing for desert areas at a rate greater than that for the selected urban cities with increase in spring season more significant than the annual scale and the summer seasons. For the urban cities, the rainfall amount is even decreasing on annual scale and summer season with a slight increase in the spring which is still less than the rate for the desert areas. The rain days are increasing for desert areas at a rate greater than the selected urban cities. The increase is more significant in the spring season than the annual and the summer seasons. Summer rain days are decreasing. Just like for the rainfall, the rain days for urban cities are decreasing on annual scale and summer season with a slight increase in the spring. This increase in spring is still less than the rate of increase for the desert areas. Considering precipitation extremes, both the desert areas and urban cities are exhibiting an increasing trend of spring precipitation extremes. The desert areas have a slightly greater rate than urban cities. For summer precipitation extreme, the rate is decreasing for the urban cities and even a smaller rate, though still increasing for the desert areas. This situation probably could imply increase in the incidence of droughts in urban cities during the summer season and therefore, there should be careful utilisation of water resources.

CHAPTER FIVE CONCLUSION AND RECOMMENDATIONS

Results from this study have generally shown that the trends in precipitation and temperature are not the same for both desert areas and urban cities which were considered for the study and over the study period. It was also evident that even the desert areas themselves or even the urban cities themselves, the trends were sometimes not the same. There were for instance cases where some stations could present increasing trends while others could present decreasing trend. This study has therefore answered the key question of this research which was: ‘are the trends the same?’ and the answer is that the trends are not the same.

Another question is therefore invited. Which areas are areas (stations) likely to present worsening scenarios in future? Will there be a blur in terms of seasons? These and many other questions need to be studied in details.

DISSERTATION SUBMITTED TO NANJING UNIVERSITY OF INFORMATION SCIENCE & TECHNOLOGY REFERENCES AND BIBLIOGRAPHY

1. Ahmed U. J. 2007: Industrialization in North-Eastern Region, A Mittal Publication, New Delhi.

2. Ahrens C. D. 2008: Essentials of Meteorology: an invitation to the atmosphere, Thomson Learning Inc., USA.

3. Akinremi O., McGinn and Cutforth H. 1999: Precipitation trends on the Canadian Prairies, J. Clim., 12, pp2996-3003.

4. Aksu H., Kuscu S. and Simsek O. 2010: Trend analysis of hydro- meteorological parameters in climate regions of Turkey, State Hydraulic works, turkey, pp1-7.

5. Anderson, T. W., 1971: The Statistical Analysis of Time Series, Wiley, New York, pp755.

6. Baas S. and Selvaraju R. 2007: Climate variability and change; adaptation to drought in Bangladesh. A resource training book. FAO, Rome, pp56.

7. Barry R. G. and Chorley R. J. 1998: Atmosphere, weather and climate, 7th edn. Routledge, New York, pp40.

8. Baudoin W.O. and Zabeltitz C.V. 1999: Greenhouses and shelter structures for tropical regions. FAO plant production and protection paper 154. FAO, Italy, pp41.

9. Bloomer J. B., Vinnikov Y. K. and Dickerson R.R. 2010: Changes in seasonal and diurnal cycles of Ozone and temperature in the eastern U.S. J. Atm. Env., 44, pp543-2551.

10. Bonsal, B. R., X. Zhang, L. A. Vincent, and W. D. Hogg, 2001: Characteristics of daily and extreme temperature over Canada. J. Climate, 14, pp1959–1976.

11. Bretherton, C. S., C. Smith, and J. M. Wallace. 1992: An inter- comparison of methods for finding coupled patterns in climate data. J. Clim., 5, pp541-560.

REFERENCES AND BIBLIOGRAPHY

12. Buda A. and DeWalle R. D. 2002: Potential effects of changes in precipitation and temperature on wet deposition in Central Pennsylvania. J. Atm. Env., 36, pp3767-3778.

13. Cox W. G. 2010: Bird Migration and global Change, Library of Congress Cataloging-in-Publication Data, USA, pp291.

14. Feng S. and Hu Q. 2004: Variations in the Teleconnnection of ENSO and Summer Rainfall in Northern China: A role of the Indian Summer Monsoon. J. Clim., 17, pp4871-4881.

15. Feuerbacher B. and Stoewer H. 2006: Utilisation of space: today and tomorrow, Springer-Verlag Berlin Heidelberg, Germany, pp123.

16. Folland C. and Anderson C., 2002: Estimating Changing Extremes Using Empirical Ranking Methods, J. Climate, 15, pp2954-2960

17. Frakes L. A., Pittock A. B. and Jenssen D. 1978: Climate Change and Variability, a Southern Perspective, New York, Cambridge University Press, pp190-197.

18. Gemmer M., Fischer T., Jiang T., Su B., and Liu L, 2010: Trends in precipitation extremes in the Zhujiang river basin, South China, J. Clim., 24, pp750-761.

19. Hanif M. 2005: Encyclopedia of Agricultural Geography, Anmol Publication pvt Ltd., New Delhi, pp142-144.

20. Harger J. R. E., 1995: Air temperature variations and ENSO effects in Indonesia, the Philippines and Salvador. ENSO patterns and changes from 1866- 1993. J. Atm. Env., 29(16), pp1919-1942.

21. Harris W. and Kay R. 1994: Weather and Climate. R.I.C Publications, Greenwood, pp33.

22. Intergovernmental Panel on Climate Change, 4th Assessment Report.

23. Jagannathan P. and Parthasarathy B. 1973: Trends and periodicities of rainfall over India, Indian Institute of Tropical Meteorology, Poona 5, India, Monthly Weather Review, 101(4), pp371 – 375.

DISSERTATION SUBMITTED TO NANJING UNIVERSITY OF INFORMATION SCIENCE & TECHNOLOGY 24. Kimei M. and Khabongo L. School certificate geography 1. East African Publishers, Nairobi, Kenya. 2008, pp35.

25. Kubota H., Shirooka R., Ushiyama T., Chuda T., Iwasaki S. and Takeuchi K. 2005: Seasonal variations of precipitation properties associated with the monsoon over Palau in the Western Pacific., J. Hydrometeo., 6, pp518-531.

26. Kusky T. 2009: Climate Change; shifting glaciers, deserts, and climate belts. Library of Cogress Cataloging-in-Publication Data, USA, pp156.

27. Laity J. 2008: Desert and desert environments, Markono Print Media, Singapore, pp1-2, pp282-289

28. Lau L. S. and Mink J. F. 2006: Hydrology of the Hawaiian Islands. Library of Congress Cataloging-in-Publication Data, USA, pp39-53.

29. Lazo K. J., Lawson M., Larsen H. P. and Waldman M. D. U.S. 2011: Economic sensitivity to weather variability, AMS, pp709-720.

30. Linacre E. 1992: Climate data and resource; a reference guide, British Library cataloguing in publication data, New York, pp37.

31. Longjun C. and Xiaohui Y. 2010: Desertification and it’s control in China. Higher Education Press, Beijing. pp29-30.

32. MacKinnon J. and Phillips K. 2006: A Field Guide to the Birds of China, China, pp5-6.

33. Maloney S. E. 2008: Chapman, the Boater’s Handbook, 4th edn. Hearst Communication, Inc. China, pp132.

34. Maslin M. Global warming; a very short introduction. Oxford University Press. 2009, Great Clarendon Street, Oxford, pp176.

35. Maxwell J. C. and Pesic P., Theory of Heat, 2001, General Publishing Company Ltd, Canada, pp1-32.

36. Meehl A., Zwiers F., Evans J., Knutson T., Mearns L., and Whetton P. 2000: Trends in Extreme Weather and Climate Events: Issues Related to Modeling Extremes in Projections of Future Climate Change. Bull. Amer. Meteor. Soc., 81(3), pp427-436.

REFERENCES AND BIBLIOGRAPHY

37. Michaelides S., Levizzani V., Anagnostou E., Bauer P., Kasparis T. and Lane J. E., 2009: Precipitation: measurement, remote sensing, climatology and modeling, J. Atm. Res., 94, pp512-533.

38. Morss E. R., Lazo K. J., Brown G. B., Brooks E. H., Ganderton T. P. and Mills N. B. 2008: Societal and economic research and applications for weather forecasts, AMS, pp335-346.

39. Murphey R. The scope of geography, 3rd ed. 1982: Library of Congress Cataloging in Publication Data.

40. Ninno del C., Dorosh A. P., Lisa C. S. and Roy K. D., 2001: The 1998 Floods in Bangladesh; disaster impacts, household coping, strategies, and response. Research report 122, Internal Food Policy Research Institute, Washington D.C.

41. Oerlemans J. 2001: Glaciers and Climate Change, Wilco b. v., Netherlands, pp52.

42. Olivier E., Etienne P., Karel V., Bas V. W., 2007: Effectiveness of erosion mitigation measures to prevent muddy floods; a case study in the Belgian loam belt; J. Agri., Ecosystems and Environment, 118, pp149-158.

43. Piringer et al. Investigating the surface energy balance in urban areas, recent advances and future needs, pp1-16.

44. Qian C., Fu C. and Wu Z. 2011: Changes in the Amplitude of the Temperature Annual Cycle in China and Their Implication for Climate Change Research. J. Climate, 24, pp5292-5302.

45. Rao P. V. 2008: Agricultural Meteorology. Prentice-Hall of India pvt. Ltd., New Delhi.

46. Rooij V. B. 2006: Regulating land and pollution in China; lawmaking, compliance, and enforcement; theory and cases, Leiden University Press, pp368.

47. Roy S. and Yuan F. 2009: Trends in Extreme Temperature in Relation to Urbanization in the Twin Cities Metropolitan Area, Minnesota, J. App. Meteo. & Clim. 48, pp669-679.

DISSERTATION SUBMITTED TO NANJING UNIVERSITY OF INFORMATION SCIENCE & TECHNOLOGY 48. Sailor J. D. 1993: Role of surface characteristics in urban meteorology and air quality, Ph.D thesis submitted to Mechanical Engineering Department of University of California, pp1-185.

49. Saucier J. W. 2003: Principles of Meteorological Analysis. Dover Publications, Inc. USA.

50. Schmal H. 1981: Patterns of European urbanization since 1500, Croom Helm Ltd, London.

51. Srivastara G. P. 2008: Surface meteorological instruments and measurement practices, Atlantic Publishers and Distributors (P) Ltd., New Delhi, pp89-96.

52. Stefan U., Steinbrich A., Tetzlaff D. and Leibundgut C. 2002: Regional analysis of the generation of extreme floods. FRIEND 2002-Regional hydrology: Bridging the gap between Research and Practice (proceedings of the 4th International FRIEND Conference held at Cape Town, South Africa, March 2002). IAHS Publ. no. 274, pp 243-249.

53. Stone D., Weaver J. A. and Stouffer J. R. 2001: Projections of climate change onto modes of atmospheric variability. AMS, J. Climate, 14, pp3551-3565.

54. US Department of Interior and US Geological Survey, Water-use trends in the desert southwest; 1950-2000, Ground Water Resource Program, Scientific investigation report.

55. US National Research Council. 2000: Reconciling Observations of Global Temperature change, National Academy Press, Washington D.C.

56. Vinnikov Y. K., Robock A., Cavalieri D. J. and Parkinson L. C. 2002: Analysis of seasonal cycles in Climatic trends with application to satellite observations of sea ice extent. Geophys. Res. Let., 29 (9), pp24-1to24-4

57. Wang C., Zhang Z. and Tian W. 2011: Factors affecting the surface radiation trends over China between 1960 and 2000. J. Atm. Environment, 45, pp2379-2385.

REFERENCES AND BIBLIOGRAPHY

58. Wang X. L. and Cho H. 1997: Spatial-Temporal structures of trend and oscillatory variabilities of precipitation over Northern Eurasia, American Meteorology Society., J. Climate, 10, pp2285-2298.

59. Wang K., Ye H., Chen F., Xiong Y. and Wang C. 2011: Urbanization effect on the diurnal temperature range: different roles under solar dimming and brightening, J. Climate. 25(3), pp1022-1027.

60. Watson R., Zinyowera C. and Moss H. 1998: The regional impacts of climate change: an assessment of vulnerability, IPCC, Cambridge university press, USA.

61. Waylen P. and Caviedes C., 1990: Annual and seasonal fluctuations of precipitation and stream flow in the Aconcagua River. J. Hydrology. 120(1-4), pp79-102.

62. Wilks S. D. 2011: Statistical methods in the atmospheric science, 3rd edition, Academic Press, Elsevier, USA, pp166-168.

63. Zhai P., Zhang X., Wan H. and Pan X., 2005: Trends in Total Precpitation and Freequency of Daily Precpitation Extremes over China. J. Climate, 18, pp1096 – 1108.

64. Zhang Q., Singh P., Sun P., Chen X., Zhang Z. and Li J. 2011: Precpitation and streamflow changes in China: Changing patterns, causes and implications, J. Hydrology, 10, pp204-216.

65. Zhou L., Dickinson E. R., Tian Y., Fang J., Li Q., Kaufmann K. R., Tucker J. C., and Myneni B. R., 2004: Evidence for a significant urbanization effect on climate in China, PNAS, 101(26), pp9540-9544.

DISSERTATION SUBMITTED TO NANJING UNIVERSITY OF INFORMATION SCIENCE & TECHNOLOGY APPENDICES

A1: Data tables – Station data Table

A1: Annual precipitation of selected urban cities

Year

Beijing Changchun Dalian Datong Harbin Hohhot Shenyang Taiyuan Tangshan Tianjin 1981 -0.26 -0.07 -0.24 0.10 0.16 0.15 -0.17 -0.19 -0.12 0.10 1982 0.03 -0.43 -0.34 -0.08 -0.01 -0.15 -0.21 0.05 0.14 -0.36 1983 -0.07 0.02 0.02 0.10 -0.01 0.07 0.09 0.18 0.05 -0.07 1984 -0.08 0.13 0.03 -0.33 0.15 0.22 0.03 -0.18 0.15 0.21 1985 0.36 0.42 0.59 0.12 0.39 0.09 0.16 0.28 0.60 0.15 1986 0.26 0.36 -0.33 -0.13 -0.05 -0.34 0.39 -0.39 0.22 0.06 1987 0.29 0.03 0.34 0.03 0.27 -0.38 0.04 -0.04 0.24 0.11 1988 0.27 -0.19 -0.21 0.20 0.13 0.07 -0.10 0.37 0.10 0.39 1989 -0.16 0.08 -0.23 0.07 -0.36 -0.09 -0.33 -0.01 -0.25 -0.36 1990 0.32 0.00 0.10 0.20 -0.09 0.20 -0.08 0.11 0.25 0.17 1991 0.41 0.21 -0.24 0.14 0.11 -0.07 0.07 0.04 0.26 0.15 1992 0.02 -0.10 0.29 0.16 -0.14 0.34 -0.21 -0.08 -0.32 -0.28 1993 0.23 -0.19 -0.20 -0.34 0.03 -0.26 -0.09 -0.09 -0.20 0.03 1994 0.36 0.19 0.37 -0.05 0.52 0.27 0.28 0.01 -0.04 0.38 1995 0.08 -0.03 0.13 0.50 -0.21 0.17 0.26 0.16 0.22 0.40 1996 0.11 -0.15 0.14 0.17 -0.11 -0.11 0.03 0.54 -0.03 -0.08 1997 -0.35 0.00 -0.09 -0.24 -0.10 -0.02 -0.18 -0.41 -0.38 -0.26 1998 0.38 0.08 0.41 0.08 0.23 0.42 0.23 -0.12 0.35 -0.02 1999 -0.50 -0.15 -0.55 -0.24 -0.18 -0.35 -0.20 -0.18 -0.30 -0.24 2000 -0.30 -0.28 -0.28 -0.22 -0.09 -0.20 -0.28 -0.01 -0.30 -0.10 2001 -0.41 -0.32 -0.16 -0.15 -0.28 -0.25 -0.16 -0.30 -0.10 -0.02 2002 -0.34 -0.15 -0.46 0.02 0.12 0.22 0.06 -0.01 -0.46 -0.25 2003 -0.24 -0.10 -0.10 0.08 -0.04 0.65 -0.07 0.24 -0.10 0.27 2004 -0.09 -0.17 0.06 0.17 -0.02 0.05 0.01 -0.11 0.08 -0.04 2005 -0.28 0.18 0.33 -0.05 -0.06 -0.37 0.18 -0.35 0.30 0.21 2006 -0.40 0.10 -0.15 -0.27 -0.09 -0.27 -0.17 0.00 -0.26 -0.19 2007 -0.09 -0.07 0.48 0.04 -0.17 -0.34 -0.04 0.27 0.03 -0.24 2008 0.11 0.24 -0.12 0.10 -0.18 0.44 0.03 -0.16 0.04 0.08 2009 -0.08 -0.17 0.24 -0.31 -0.01 -0.33 -0.06 0.48 -0.10 0.11 2010 0.41 0.52 0.18 0.13 0.10 0.18 0.48 -0.11 -0.07 -0.31 30yr Avg 529.5 577.1 579.8 369.4 537.7 396.4 698.5 423.2 590.1 511.5

APPENDICES

Table

A2: Annual precipitation of selected desert areas

Qaidam

Year -

Alza Zuoqi Bayan MOD Da Guaizihu Hami Hoboksar Linhe Qiemo Ruoqiang Turpan 1981 171.3 85.0 117.9 87.8 42.4 148.9 140.6 40.6 106.9 13.9 1982 122.3 37.5 101.6 15.6 34.6 85.0 98.5 38.9 30.7 3.8 1983 211.4 81.7 100.8 51.3 32.8 179.8 130.7 12.0 18.7 4.9 1984 244.3 101.1 57.7 20.8 51.1 135.1 207.9 30.7 40.8 30.2 1985 195.5 49.9 42.5 19.6 17.2 90.3 121.1 32.5 15.3 8.1 1986 195.2 63.6 101.1 16.2 24.1 153.5 55.9 5.9 10.9 8.6 1987 190.7 60.6 87.8 13.9 47.5 118.0 139.7 37.4 24.4 26.7 1988 236.2 92.7 108.6 33.2 68.3 148.9 267.9 46.1 64.3 27.0 1989 144.3 79.2 102.7 28.0 29.7 176.6 126.8 42.2 19.1 20.9 1990 215.7 50.6 49.5 43.9 31.3 134.7 157.6 18.4 18.9 16.3 1991 183.3 76.0 121.8 40.9 20.1 91.5 117.8 26.2 22.5 8.5 1992 261.9 60.5 147.2 54.5 71.7 208.5 190.1 18.4 40.8 23.2 1993 171.7 120.5 91.1 18.5 47.9 255.6 89.7 13.5 45.9 7.2 1994 234.4 135.0 56.9 40.1 29.1 125.2 176.4 17.8 9.3 21.3 1995 241.2 136.2 44.8 71.5 62.3 133.2 243.1 22.4 29.4 11.4 1996 275.1 167.5 62.3 78.4 28.6 107.0 146.5 34.9 23.5 10.4 1997 202.8 74.5 61.9 59.8 9.2 71.9 213.5 21.8 13.9 5.5 1998 279.3 175.1 87.9 28.1 65.9 134.3 180.4 21.1 60.5 33.4 1999 149.2 77.3 75.8 64.7 36.4 153.0 114.4 19.3 29.7 9.9 2000 187.4 98.7 53.9 24.3 53.1 141.4 96.0 45.3 38.2 16.4 2001 225.3 67.1 49.2 25.5 55.5 175.5 159.0 7.0 5.1 16.7 2002 257.1 150.2 164.9 39.2 68.2 188.5 148.4 36.2 40.1 25.6 2003 330.1 147.9 114.2 49.6 57.4 147.6 140.3 38.0 56.9 30.9 2004 208.4 113.1 88.1 31.3 37.5 164.9 160.6 10.4 20.1 10.5 2005 126.2 101.7 95.2 14.5 66.1 168.2 77.0 43.4 117.0 9.0 2006 144.5 121.4 115.3 21.5 29.2 80.1 175.4 26.9 40.3 8.2 2007 246.0 145.0 124.8 81.3 55.7 285.4 165.8 54.3 68.3 12.3 2008 291.2 120.6 115.9 62.2 42.1 92.6 200.9 8.5 24.7 23.2 2009 134.8 119.2 100.6 33.4 31.2 112.1 87.9 17.8 23.1 6.9 2010 159.4 145.6 142.8 66.8 64.9 209.9 136.2 37.9 57.8 7.0 30yr Avg 207.9 101.8 92.8 41.2 43.7 147.2 148.9 27.5 37.2 15.3

DISSERTATION SUBMITTED TO NANJING UNIVERSITY OF INFORMATION SCIENCE & TECHNOLOGY

Table A3: Annual number of rain days of selected urban cities

Year

Beijing Changchun Dalian Datong Harbin Hohhot Shenyang Taiyuan Tangshan Tianjin 1981 65 93 71 65 104 69 85 54 66 56 1982 64 76 61 79 96 72 80 71 59 56 1983 63 98 71 74 104 69 73 84 58 55 1984 69 90 76 66 123 74 77 66 64 54 1985 89 105 92 69 116 77 95 70 92 76 1986 67 110 74 72 103 54 105 63 77 67 1987 74 100 73 69 117 59 93 66 77 72 1988 77 99 61 90 119 78 77 75 64 66 1989 60 88 65 70 103 71 75 63 71 55 1990 87 115 91 82 102 72 106 95 89 83 1991 84 115 71 78 105 67 98 67 79 73 1992 81 103 68 85 96 81 90 70 57 59 1993 112 100 68 63 107 65 95 66 69 62 1994 63 105 66 67 105 72 95 66 57 62 1995 68 93 77 66 101 60 84 60 68 58 1996 74 89 75 65 98 76 84 82 60 59 1997 54 85 54 60 93 58 69 43 47 53 1998 82 93 70 72 98 86 89 63 68 66 1999 63 99 54 59 99 53 75 48 55 52 2000 61 103 71 72 104 61 87 67 61 56 2001 68 83 72 72 77 60 80 71 72 62 2002 61 98 55 77 105 69 75 72 57 56 2003 74 92 76 98 91 101 75 93 67 59 2004 65 93 69 74 95 79 87 69 64 56 2005 63 106 72 62 106 58 88 58 64 61 2006 59 90 72 71 83 61 90 66 57 54 2007 57 79 85 69 86 65 81 77 61 63 2008 75 104 71 82 91 73 75 74 74 63 2009 65 105 64 67 108 62 91 70 67 62 2010 72 122 80 70 119 69 121 67 63 57 30yr Avg 71 98 71 72 102 69 87 69 66 61

APPENDICES

Table A4: Annual number of rain days of selected desert areas

Qaidam

Year -

Alza Zuoqi Bayan MOD Da Guaizihu Hami Hoboksar Linhe Qiemo Ruoqiang Turpan 1981 48 38 40 20 24 73 36 14 17 13 1982 48 32 45 15 20 70 36 11 18 7 1983 55 37 46 26 26 65 39 14 16 12 1984 57 36 22 23 23 93 38 11 12 9 1985 50 21 25 15 15 71 32 8 7 5 1986 52 27 37 18 22 77 32 6 11 12 1987 51 30 44 16 37 75 36 13 15 16 1988 56 33 39 20 29 84 42 11 18 18 1989 53 28 49 17 26 76 34 23 13 17 1990 49 36 34 17 30 72 52 11 22 16 1991 45 30 39 13 18 63 46 14 11 8 1992 62 36 41 20 28 95 43 13 13 9 1993 48 33 45 14 31 104 30 18 13 8 1994 65 44 28 19 19 90 44 17 11 12 1995 50 34 29 23 23 70 41 17 13 13 1996 59 45 25 21 22 70 39 14 15 8 1997 40 38 28 18 18 53 25 9 13 8 1998 52 50 38 18 39 82 40 15 11 17 1999 44 28 35 17 27 82 28 9 8 13 2000 55 37 34 12 39 81 32 19 16 14 2001 45 27 25 15 25 83 26 7 10 9 2002 52 40 43 19 23 80 36 23 11 18 2003 62 45 45 21 33 90 50 15 18 24 2004 54 43 37 19 25 87 40 9 9 22 2005 37 34 33 13 24 69 24 17 15 14 2006 51 39 46 19 35 74 35 19 24 14 2007 62 40 44 23 28 85 37 10 13 17 2008 59 40 47 20 29 55 47 7 12 13 2009 43 28 41 14 19 83 22 10 15 12 2010 51 35 55 21 28 96 35 21 22 15 30yr Avg 52 35 38 18 26 78 37 14 14 13

DISSERTATION SUBMITTED TO NANJING UNIVERSITY OF INFORMATION SCIENCE & TECHNOLOGY

Table A5: Annual maximum temperature for selected urban cities

Year

Beijing Changchun Dalian Datong Harbin Hohhot Shenyang Taiyuan Tangshan Tianjin 1981 18.1 10.4 14.5 13.3 9.7 12.6 14.1 16.8 17.9 18.3 1982 18.8 12.4 15.5 14.4 11.6 13.6 15.0 17.5 17.8 19.0 1983 19.0 11.7 15.5 14.0 10.1 13.5 14.7 16.7 17.9 19.2 1984 17.4 10.6 14.0 13.0 9.3 12.1 13.6 15.8 16.5 17.4 1985 16.8 10.3 13.2 13.2 9.2 12.7 13.1 16.0 15.9 16.8 1986 17.8 10.3 14.2 13.5 9.6 13.0 13.4 16.9 16.7 17.8 1987 17.8 10.5 14.1 14.9 9.2 14.0 13.7 17.7 16.6 17.9 1988 18.1 11.3 14.8 13.4 9.8 12.8 14.5 16.7 17.2 18.1 1989 18.7 12.1 15.5 14.0 11.1 13.9 15.0 16.8 17.9 18.9 1990 18.0 12.0 14.8 14.4 11.2 14.1 14.4 17.1 17.1 18.1 1991 17.7 10.8 14.9 13.9 10.4 13.8 13.6 17.1 17.1 17.9 1992 18.2 11.3 15.1 13.6 10.4 13.5 14.0 17.0 17.6 18.1 1993 18.3 11.3 14.9 13.7 9.9 13.1 14.0 16.6 17.6 18.1 1994 19.0 11.7 15.5 14.6 10.3 14.0 14.7 17.6 18.2 18.8 1995 18.7 11.9 14.9 14.0 11.0 13.4 14.5 17.5 17.7 18.5 1996 17.9 11.3 14.5 13.6 10.7 12.9 13.9 16.7 17.1 17.6 1997 18.5 12.1 15.6 15.2 11.2 14.4 14.9 18.3 18.2 18.7 1998 18.5 12.6 15.5 15.7 11.5 14.7 15.3 18.9 18.3 19.1 1999 18.7 11.5 15.9 15.8 10.2 14.9 14.6 18.7 18.2 19.0 2000 18.3 10.9 15.1 14.4 10.0 13.8 13.8 17.3 17.9 18.5 2001 18.3 11.5 15.4 15.0 10.4 14.6 14.2 17.7 18.1 18.7 2002 18.7 12.3 15.5 14.8 10.9 14.2 15.0 18.0 18.3 18.9 2003 18.1 12.3 14.9 13.6 11.4 12.8 14.6 16.6 17.8 18.2 2004 18.8 12.5 15.8 14.5 11.1 14.2 15.1 17.8 18.4 18.9 2005 18.5 10.7 14.5 14.2 9.7 13.9 13.6 17.8 18.0 18.6 2006 18.7 11.9 15.1 15.4 10.7 15.0 14.4 18.4 18.2 18.7 2007 19.3 13.0 15.8 15.5 12.0 15.1 15.2 18.0 18.6 19.1 2008 18.7 12.5 14.9 14.4 12.0 13.4 14.9 17.6 17.8 18.8 2009 18.7 11.7 15.3 15.0 10.3 14.4 14.1 17.7 17.7 18.6 2010 17.6 10.0 13.7 14.4 9.6 13.8 12.8 17.7 16.8 17.5 30yr Avg 18.3 11.5 15.0 14.3 10.5 13.7 14.3 17.4 17.6 18.4

APPENDICES

Table A6: Annual minimum temperature for selected urban cities

Year aiyuan

Beijing Changchun Dalian Datong Harbin Hohhot Shenyang T Tangshan Tianjin 1981 7.2 0.0 7.4 0.1 -1.6 0.2 3.2 3.6 6.6 8.3 1982 7.6 1.2 8.2 0.7 -0.9 1.0 3.9 4.0 5.9 8.9 1983 7.9 0.9 8.4 0.7 -1.4 0.9 4.1 3.8 6.0 9.1 1984 6.8 -0.1 7.7 -0.7 -2.3 -0.5 3.4 3.1 5.0 8.2 1985 6.8 0.2 7.0 0.1 -2.4 0.3 3.4 3.9 5.2 8.2 1986 7.0 0.5 7.3 -0.3 -1.6 0.2 3.7 3.2 5.4 8.5 1987 7.4 0.4 7.2 1.0 -2.3 1.2 4.0 3.8 5.9 8.6 1988 7.9 0.9 8.2 0.4 -1.4 0.8 4.6 3.8 5.8 8.8 1989 8.2 1.8 8.7 1.3 -1.0 1.7 3.0 4.3 6.0 9.2 1990 8.1 2.5 8.4 1.3 0.0 1.8 3.7 4.2 6.5 9.2 1991 7.8 1.2 8.1 0.7 -1.3 1.4 2.9 3.9 5.8 8.7 1992 7.9 1.2 8.3 0.8 -1.8 1.1 2.7 3.6 5.8 7.5 1993 8.0 1.3 8.3 0.3 -1.4 0.4 2.7 3.3 5.6 7.3 1994 8.9 2.0 8.9 1.0 -1.0 1.8 3.7 4.5 6.8 8.3 1995 8.5 1.9 8.5 0.6 -0.2 0.9 3.3 3.9 6.2 8.2 1996 8.2 1.1 8.1 0.6 -0.5 0.7 2.7 4.2 7.0 7.8 1997 8.0 1.8 8.9 1.2 0.4 1.5 3.4 4.7 7.9 8.1 1998 8.4 2.6 9.0 2.7 0.9 2.6 4.7 5.3 8.6 8.9 1999 8.0 1.3 9.1 2.4 -0.2 2.4 4.1 5.4 8.3 8.7 2000 7.7 0.9 8.2 1.2 -0.8 1.7 3.1 5.0 8.2 8.4 2001 7.8 1.4 8.4 1.7 -0.4 2.5 3.1 5.1 8.2 8.8 2002 8.0 2.1 8.9 2.1 0.5 2.5 4.3 5.4 8.6 8.8 2003 8.2 2.2 8.6 1.5 1.2 2.2 4.3 5.0 8.7 8.8 2004 8.9 2.4 9.3 2.0 0.9 2.7 4.7 5.2 8.7 8.7 2005 8.6 1.4 8.0 1.2 0.2 2.1 3.3 5.2 8.3 8.2 2006 8.9 1.9 8.6 2.3 0.2 3.2 2.8 6.4 8.8 8.9 2007 9.4 2.9 9.5 2.4 1.7 3.5 3.5 6.0 9.0 9.3 2008 8.7 2.6 8.6 1.4 1.6 2.1 3.0 5.4 6.2 8.8 2009 8.5 1.2 8.5 1.8 0.0 2.3 2.0 5.6 6.2 8.4 2010 8.2 0.8 7.4 2.2 -0.3 2.2 2.2 6.0 5.4 8.0 30yr Avg 8.1 1.4 8.3 1.2 -0.5 1.6 3.4 4.6 6.9 8.5

DISSERTATION SUBMITTED TO NANJING UNIVERSITY OF INFORMATION SCIENCE & TECHNOLOGY

Table A7: Annual DTR for selected urban cities

Year ijing

Be Changchun Dalian Datong Harbin Hohhot Shenyang Taiyuan Tangshan Tianjin 1981 10.9 10.4 7.1 13.2 11.4 12.3 10.9 13.2 11.3 10.0 1982 11.1 11.2 7.3 13.6 12.5 12.6 11.1 13.4 11.9 10.0 1983 11.1 10.8 7.1 13.3 11.5 12.6 10.6 12.9 11.9 10.1 1984 10.6 10.6 6.3 13.7 11.6 12.7 10.2 12.6 11.5 9.3 1985 9.9 10.1 6.2 13.1 11.5 12.4 9.7 12.2 10.7 8.7 1986 10.7 9.8 6.9 13.8 11.2 12.8 9.7 13.6 11.2 9.3 1987 10.4 10.1 6.9 14.0 11.4 12.9 9.7 13.8 10.7 9.3 1988 10.2 10.4 6.6 13.0 11.2 12.0 9.9 12.9 11.4 9.3 1989 10.4 10.3 6.8 12.7 12.1 12.2 12.0 12.6 11.9 9.7 1990 9.9 9.5 6.4 13.0 11.2 12.2 10.7 12.9 10.5 9.0 1991 9.9 9.6 6.8 13.2 11.6 12.4 10.7 13.2 11.3 9.2 1992 10.3 10.2 6.8 12.8 12.1 12.4 11.3 13.4 11.7 10.6 1993 10.3 10.0 6.6 13.3 11.3 12.6 11.3 13.3 11.9 10.8 1994 10.1 9.6 6.6 13.6 11.4 12.3 11.1 13.1 11.4 10.5 1995 10.2 10.0 6.4 13.4 11.3 12.5 11.1 13.6 11.6 10.3 1996 9.7 10.1 6.4 13.0 11.1 12.3 11.2 12.5 10.1 9.8 1997 10.5 10.2 6.7 13.9 10.9 13.0 11.5 13.7 10.3 10.6 1998 10.1 9.9 6.5 13.0 10.6 12.1 10.6 13.6 9.7 10.2 1999 10.7 10.2 6.8 13.5 10.4 12.5 10.5 13.3 9.9 10.3 2000 10.6 10.0 6.9 13.2 10.8 12.2 10.7 12.4 9.7 10.2 2001 10.5 10.1 6.9 13.3 10.8 12.0 11.1 12.6 9.9 9.8 2002 10.7 10.1 6.6 12.7 10.4 11.7 10.7 12.6 9.7 10.1 2003 9.8 10.0 6.3 12.1 10.2 10.7 10.4 11.6 9.2 9.3 2004 9.8 10.1 6.6 12.6 10.2 11.6 10.4 12.7 9.8 10.2 2005 10.0 9.3 6.4 13.1 9.6 11.8 10.3 12.6 9.7 10.4 2006 9.8 10.0 6.6 13.2 10.6 11.8 11.6 12.0 9.4 9.8 2007 9.9 10.0 6.2 13.1 10.3 11.6 11.7 12.0 9.6 9.9 2008 9.9 9.9 6.3 13.0 10.4 11.3 12.0 12.2 11.6 10.0 2009 10.3 10.5 6.8 13.2 10.4 12.2 12.1 12.1 11.5 10.2 2010 9.4 9.2 6.3 12.2 9.9 11.5 10.6 11.7 11.3 9.5 30yr Avg 10.3 10.1 6.6 13.2 11.0 12.2 10.8 12.8 10.7 9.9

APPENDICES

Table A8: Annual maximum temperature for selected desert areas

Qaidam

Year -

Alza Zuoqi Bayan MOD Da Guaizihu Hami Hoboksar Linhe Qiemo Ruoqiang Turpan 1981 13.7 14.1 9.6 16.2 17.4 10.0 14.7 18.1 19.4 21.1 1982 14.4 15.1 9.2 17.4 18.7 11.1 15.4 18.4 19.8 22.1 1983 13.6 14.4 8.4 16.8 18.1 10.9 14.9 18.4 19.6 21.9 1984 13.0 13.3 9.0 15.5 17.1 8.3 13.6 17.9 18.8 20.3 1985 13.6 13.9 9.5 16.3 17.8 9.7 14.3 18.8 20.0 20.9 1986 13.6 14.1 9.3 16.6 18.0 10.1 14.6 18.6 20.0 21.6 1987 15.1 15.4 10.3 17.6 17.3 9.8 15.8 19.0 20.2 21.3 1988 13.8 14.2 9.9 16.6 17.6 9.7 14.6 18.7 19.9 21.3 1989 14.1 14.7 9.4 17.3 18.6 10.3 15.1 18.4 20.1 22.1 1990 14.9 15.3 10.2 17.6 18.3 10.9 15.6 19.7 21.0 22.0 1991 14.6 14.8 9.9 17.1 18.4 10.9 15.4 18.7 20.3 21.9 1992 13.9 14.4 9.0 16.8 17.5 10.0 15.1 18.4 19.4 21.0 1993 13.8 14.5 9.1 16.5 17.8 8.9 14.5 18.5 19.8 21.5 1994 14.7 15.4 10.3 17.6 18.7 9.9 15.4 19.0 20.6 21.9 1995 13.9 14.8 9.5 17.1 18.0 11.3 15.3 18.2 20.1 21.7 1996 13.8 14.4 10.0 16.7 17.6 9.7 14.6 18.6 19.8 21.0 1997 15.2 16.1 10.5 17.9 19.4 12.2 16.1 19.9 21.3 22.7 1998 15.9 16.6 11.3 18.5 18.8 11.2 16.7 19.9 21.1 22.1 1999 15.5 16.4 11.0 18.2 19.1 11.0 16.5 19.8 20.8 22.5 2000 14.9 15.5 10.2 17.4 18.2 10.6 15.3 19.3 20.3 22.1 2001 14.9 15.8 10.8 17.9 18.6 10.6 15.7 20.1 20.9 22.5 2002 15.0 15.8 10.6 17.7 19.0 11.2 15.8 19.7 21.0 22.6 2003 14.5 15.2 10.6 17.0 17.2 10.2 15.0 19.5 20.1 21.2 2004 14.8 15.5 10.4 17.5 18.5 11.0 15.5 20.2 20.9 22.0 2005 14.5 15.2 10.6 17.4 18.7 10.0 15.2 19.3 20.7 22.3 2006 15.3 15.9 11.2 17.4 19.0 11.3 16.0 19.7 21.0 22.7 2007 14.6 15.7 11.0 17.2 20.0 11.4 15.9 20.5 21.9 23.3 2008 14.1 15.2 10.3 17.4 19.5 11.5 15.3 19.2 21.1 23.2 2009 15.2 15.8 11.1 17.6 19.3 10.5 15.9 20.1 21.5 23.0 2010 14.5 15.2 11.2 17.0 18.6 9.8 15.2 19.6 20.7 22.4 30yr Avg 14.4 15.1 10.1 17.2 18.4 10.5 15.3 19.1 20.4 21.9

DISSERTATION SUBMITTED TO NANJING UNIVERSITY OF INFORMATION SCIENCE & TECHNOLOGY

Table A9: Annual minimum temperature for selected desert areas

Qaidam

Year -

Alza Zuoqi Bayan MOD Da Guaizihu Hami Hoboksar Linhe Qiemo Ruoqiang Turpan 1981 2.8 0.8 -4.6 2.1 3.1 -1.8 1.9 3.1 3.9 7.9 1982 3.1 1.5 -4.7 3.1 3.9 -0.6 2.5 3.5 4.8 8.8 1983 2.8 0.9 -6.4 2.2 3.9 -0.9 2.2 2.8 4.3 8.2 1984 2.2 0.0 -5.9 1.4 2.5 -3.2 1.2 2.3 3.3 7.4 1985 3.1 0.5 -5.6 2.0 3.1 -2.2 1.7 2.8 4.0 8.0 1986 3.1 0.8 -6.0 1.8 3.5 -1.2 1.6 3.0 4.2 9.0 1987 4.3 1.9 -4.5 3.5 3.5 -1.2 2.8 3.7 4.7 9.4 1988 3.4 1.2 -4.8 2.1 3.5 -1.8 2.4 3.3 4.6 9.3 1989 3.8 1.9 -5.3 2.8 3.4 -0.9 3.1 3.0 4.4 9.5 1990 4.0 1.8 -5.8 2.9 3.0 -0.5 3.1 3.1 4.3 9.3 1991 4.2 1.6 -5.3 2.8 2.8 -0.9 3.0 2.9 4.1 9.2 1992 3.5 1.3 -5.5 2.3 2.3 -1.1 2.8 2.9 4.0 8.5 1993 3.0 -0.2 -5.2 1.9 2.6 -1.8 2.2 3.2 4.3 8.9 1994 4.2 0.9 -5.2 3.0 2.7 -1.0 3.7 3.4 4.9 9.0 1995 3.1 -0.4 -6.4 2.9 2.4 -0.3 3.0 2.9 4.4 8.7 1996 3.3 -0.3 -5.2 2.6 2.2 -1.5 2.8 2.8 3.9 8.4 1997 4.5 0.2 -5.7 3.1 3.4 0.3 3.5 3.1 4.4 9.2 1998 4.9 1.8 -4.4 4.3 4.0 -0.3 4.8 4.3 5.5 9.6 1999 4.8 1.2 -4.6 3.8 3.9 -0.3 4.4 4.0 4.8 10.2 2000 4.2 0.7 -5.6 3.0 3.0 -1.1 3.1 3.7 4.6 9.4 2001 4.4 1.0 -5.0 3.5 3.4 -0.9 3.8 4.1 4.9 10.2 2002 4.9 1.2 -4.2 3.9 3.8 -0.1 4.1 4.2 5.1 10.8 2003 4.3 0.6 -4.2 3.0 2.7 -0.9 3.3 4.1 4.8 9.6 2004 4.3 0.8 -3.7 3.2 3.5 -0.5 3.8 4.0 5.7 9.9 2005 4.3 0.5 -3.6 2.9 3.5 -1.0 3.4 3.5 5.2 10.0 2006 5.1 1.6 -3.2 3.3 3.9 0.0 4.5 4.1 5.7 10.4 2007 4.7 1.7 -3.6 3.5 4.3 0.2 4.7 4.0 5.7 10.8 2008 3.7 0.7 -4.4 3.4 3.8 0.2 3.7 3.0 5.0 10.5 2009 4.9 1.1 -3.3 3.4 3.6 -0.4 4.0 4.0 5.0 10.5 2010 4.4 1.9 -3.0 3.7 3.7 -1.1 2.1 4.4 5.3 10.3 30yr Avg 3.9 1.0 -4.8 2.9 3.3 -0.9 3.1 3.4 4.7 9.4

APPENDICES

Table A10: Annual DTR for selected desert areas

Qaidam

Year -

Alza Zuoqi Bayan MOD Da Guaizihu Hami Hoboksar Linhe Qiemo Ruoqiang Turpan 1981 10.9 13.3 14.2 14.1 14.3 11.8 12.8 14.9 15.4 13.1 11.3 13.6 13.8 14.3 14.7 11.7 12.9 14.9 15.0 13.3 10.7 13.4 14.8 14.5 14.2 11.7 12.7 15.5 15.3 13.7 10.7 13.3 14.9 14.1 14.5 11.5 12.5 15.6 15.5 12.9 1985 10.5 13.3 15.1 14.2 14.7 11.9 12.5 15.9 16.0 12.9 10.5 13.2 15.2 14.7 14.5 11.3 13.0 15.6 15.7 12.6 10.8 13.4 14.8 14.1 13.8 11.0 13.0 15.3 15.5 11.8 10.4 13.0 14.7 14.5 14.1 11.4 12.1 15.4 15.3 12.0 1989 10.3 12.8 14.7 14.4 15.2 11.1 11.9 15.4 15.6 12.5 10.9 13.5 16.0 14.6 15.3 11.4 12.4 16.5 16.6 12.7 10.4 13.2 15.3 14.3 15.6 11.8 12.4 15.8 16.2 12.7 10.4 13.2 14.5 14.5 15.2 11.1 12.3 15.5 15.4 12.5 1993 10.9 14.7 14.3 14.6 15.2 10.7 12.3 15.2 15.5 12.6 10.6 14.6 15.5 14.6 15.9 10.9 11.8 15.7 15.7 12.9 10.7 15.2 15.9 14.1 15.6 11.7 12.4 15.3 15.6 13.0 10.5 14.7 15.2 14.1 15.4 11.2 11.8 15.8 15.9 12.6 1997 10.7 15.9 16.1 14.7 16.0 11.9 12.6 16.8 16.9 13.5 11.0 14.8 15.7 14.1 14.8 11.4 11.9 15.5 15.6 12.4 10.6 15.2 15.6 14.4 15.1 11.3 12.1 15.8 16.0 12.3 10.7 14.8 15.8 14.4 15.2 11.8 12.3 15.6 15.6 12.7 2001 10.5 14.7 15.8 14.4 15.2 11.5 11.9 16.0 16.0 12.3 10.2 14.6 14.8 13.8 15.2 11.4 11.7 15.5 15.9 11.8 10.2 14.6 14.8 14.0 14.5 11.1 11.7 15.4 15.3 11.6 10.5 14.7 14.1 14.3 15.0 11.5 11.7 16.1 15.2 12.1 2005 10.3 14.7 14.2 14.6 15.2 11.1 11.8 15.8 15.5 12.2 10.2 14.4 14.3 14.1 15.1 11.4 11.5 15.6 15.3 12.2 9.9 14.0 14.6 13.8 15.7 11.2 11.2 16.4 16.2 12.5 10.4 14.5 14.7 14.0 15.6 11.3 11.6 16.2 16.1 12.7 2009 10.3 14.7 14.4 14.2 15.7 10.8 11.9 16.1 16.6 12.5 2010 10.1 13.3 14.2 13.3 14.9 10.9 13.1 15.2 15.4 12.2 30yr Avg 10.5 14.1 14.9 14.3 15.0 11.4 12.2 15.7 15.7 12.6

DISSERTATION SUBMITTED TO NANJING UNIVERSITY OF INFORMATION SCIENCE & TECHNOLOGY ACKNOWLEDGEMENT

I would like to express my gratitude to all my classmates and Ogwang Bob Alex (a current PhD student and my country mate) as well as all those who helped me including encouragement up to the completion of this piece of work. I am very grateful to NUIST School of Atmospheric Science which gave me the complete set of data that was required for the study. Particular thanks go to Ms. Xie Xiao who enabled me interpret the data in the format in which it had been archived and who offered her availability for assistance.

I would also like to thank my fellow Christian brothers and sisters from NUIST who kept on praying and encouraging me while studying this M.Sc. in Meteorology. I am greatly indebted to my family that endured the experience of missing me while I was away from them for this long period of time.

Distinguished appreciation goes to my supervisor, Prof. Shen Shuanghe who guided me especially at the starting of this study, coming up with the study topic and was always ready to help me through. Dr. Tao Sulin has been of unmatched assistance to me. I sincerely appreciate his invaluable assistance with close collaboration of Prof. Shen Shuanghe that they made me finish this study far early, ahead of time. Lastly I thank the Almighty God as I am aware that without his blessings all can be in vain. May God bless everyone who gave a helping hand in this study.