1
Simulations of water resource environment changes in Eastern China due to variations of LULC and CO2level during the last 130 years by a regional climate model*
Zheng Yiqun1+,2(郑益群),LiJunsong1 (李俊松), Zeng Xinmin1(曾新民), Qiang Xuemin1 (强学民)
1Institute of Meteorology, PLAUniversity of Sciences and Technology, Nanjing 211101, China
2Nanjing Institute of Geography and Limnology, ChineseAcademy of Sciences, Nanjing 210008, China
Submitted October 8, 2010
*Supported by the National Program on Key Basic Research Project of China (973) under Grant No. 2010CB428505, and the National Natural Science Foundation of China under Grant No. 40875067.
+ Corresponding author: . Tel: 13851694216
ABSTRACT
Anthropogenic influence on regional climate and water resource over East Asia is simulated by using of regional model nested to a global model, considering the changes of land use/land cover (LULC) and CO2 level.The results show that:variations of LULC and CO2level during the last 130 yrsmakes many regions of East Asia showing a warming trend. The most remarkable temperature increase occurs in the Inner-Mongolia, northeast and northChina, while temperature decrease in Gansu Province and north of SichuanProvince. LULC and CO2level Changesover the past 130 yrsresult in a decreasing trend of precipitation in some regions, especially in the Huai River Valley, Shandong Byland and the Yunnan-Guizhou plateau. But precipitation increase along the middle reaches of the Yangtze River,the middle reaches of the Yellow River and some areas of south China.This pattern of precipitation change with changes in surface evapotranspiration maycause a more severe drought in the lower reaches of the Yellow River and the HuaiRiver valley, the drought trend weakened at the mid- and upper reaches of the Yellow River valley, while theYangtze River valleyfloods areincreasing. Changes in LULC and CO2levelduring the last 130 yrs lead to adjustmentsin the East Asian monsoon circulation, and further affect the water vapor transportation and the water vapor budget, makingnorth Chinapresents a warm and dry climate, while Sichuan basin tends to be cold and wet, andeast China warm and wet.
Key words: anthropogenic influence, regional climate simulation, water resources environments
Citation: Zheng Yiqun, Li Junsong, Zeng Xinmin, et al., 2011:Simulations of water resource environment changes in Eastern China due to variations of LULC and CO2level during the last 130 years by a regional climate model.Acta Meteor. Sinica,XX(x), XXX-XXX.
近130年来LULC和CO2含量变化对中国东部水资源环境
影响的区域气候模拟研究
摘 要
利用区域与全球模式嵌套的气候模拟系统，通过引入现代植被利用、植被覆盖(LULC ) 和工业革命前LULC，及相应时期的CO2含量，模拟研究人类活动对东亚气候和水资源环境的影响。结果表明：近130年来LULC和CO2含量变化使得东亚许多区域呈现出升温趋势。其中，内蒙、东北、华北升温较为明显，但甘肃及四川北部地区存在降温现象。近130年来LULC和CO2含量变化使得东亚部分地区出现降水减少趋势，特别是淮河流域、山东半岛和云贵高原地区，但长江中游及华南部分地区的降水有所增加。这种降水变化格局加上地表蒸散的变化使得黄河下游及淮河流域的干旱化进一步加剧，黄河中上游地区的干旱有所减缓，而长江流域的洪涝有增加趋势。近130年来LULC和CO2含量变化导致东亚季风环流出现调整，并进一步影响到东亚水汽输送和收支状况，使得华北出现暖干、四川盆地出现冷湿，而华东地区出现暖湿化趋势。
关键词：人类活动，区域气候模拟，水资源环境
1. Introduction
Since the inception of the earth, global climate and environment changehas never been interrupted. At present, key points needed to be knownfor us are those changes that occurred in thepast centuryand strongly affected by human activities. These may also be the foundation of predicting the futurechanges (Wang et al., 2002). In recent 100 years,the climate is undergoing a significant global warming, and the warming is thought to be caused together by natural climate oscillations and human activities. According to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change (IPCC AR4),the temperature increase (1850 – 1899 to 2001 – 2005) is about 0.76 [0.57 to0.95]°C (IPCC, 2007). New evidences show that warming in the 20th century different evidently from that of natural-drivenchanges, especially warming of 0.1°C per decade during the last 50 years, might be mostly attributed to human activities (Watson et al., 2001; Qin et al., 2005). Moreover, the increase of greenhouse gases and aerosols maybe the major factors of dramatic warming since 1980’s. In fact, the radiation force produced by uniformly mixed greenhouse gases has reached +2.30[+2.07 to +2.53] W m-2in recent 200 years,which might be the main cause of the global warming (Wang et al., 2002; IPCC, 2007). The climate simulation experiments for 1860~2000 released by the IPCC report also show that, when the natural factors (Solar radiation and volcanic activities) and the artificial factors (greenhouse gases and aerosols) are synthesized, the temperature change simulations for the late 20th century are more closeto the observation data than those of including only single factor (Watson et al., 2001; Hegerlet al., 2007).These results indicate that the warming of this period are really relatedto human activities.
Anthropogenic forcinginduced global warming will enhance evaporation and may lead to more maximum intensityofprecipitation (Willett et al., 2007). But mid-high latitudes temperatureincreases larger than low latitudes in east Asian, will depress the gradient of zonal temperature and southwesterly wind intensity, and decrease the water vapor transportation from ocean to land in mid latitude, while land evaporation will get somewhat enhanced due to temperature rising. Both of the two mechanisms will cause a drought trend in this area (Zhang et al., 1999). China is one of the serious water shortage country, wherethemeanindividualwater resource only reaches to one-forth of world mean level, and the spatial-temporal distribution ofthe water resource is extremely uneven too. There are many sensitive and fragile areas of water resource in China, where the semi-wet clime regions in HuaiRiver and its north are the most sensible and fragile areas of China (Liu et al., 1996). At present, water resource environmentsof China are quite rigorous, focusing on three aspects, i.e., the serious contradiction between water supply and requirements, the accelerated deterioration water quality and more frequent hazards of floods and droughts (Song et al., 2000).
In many works, global climate models are used to evaluate how climate affected byhuman activities. By comparing numerical experiments of climate background and simulations including human activities (mainly focus on changes of greenhouse gases or land use), scientists can study climate changes affected by human activities (Watson et al., 2001; Guo et al., 2001; Ding et al., 2002; Gao et al., 2003; Fu et al., 2003; Hansen et al., 2007; Tett et al., 2007; Shindell et al., 2008; Zhang et al., 2009; O’ishi et al., 2009). However, the capability of simulating regional climates is restricted by the lower resolution of the global models. And the global models can not well reflect the influence of vegetation feed-back on regional scalesalso(Gao et al., 2006; Li et al., 2010). A rather higher resolution is needed in simulating water processes involving vegetation changes (Leung et al., 1995; Cherchi, et al., 2007). Regional climate model (RCM) has a well capability of simulatingmultiple scale interactions, and thus of delineating the features of climate changes in a regional scale(Zheng et. al.,2002a, 2003a; Tang et al., 2003; Jiao et al., 2007). They can also produce reasonable climate response to changes of underlying surfaces (Chen et al., 2001;Zheng et al., 2002b;Zheng et al., 2004; Qian et al. 2003; Cha et al., 2008; Steiner et al., 2009; Anders et al., 2009; Chang et al., 2009), and better explain the internal mechanisms of water cycle.Compared to GCM, the RCM’s finerresolution gives a morereal depiction of thesurface forcings such as topography, coastlines, inlandwater and land-surface characteristics, etc. (Christensen et al.,2007;Diaconescu et al., 2007).
In this work, by using a high resolution non-hydrostatic mesoscale model (MM5v3) which is coupled with land surface processes, we attempt to simulate the climate and water recycling processes changes affected by human activitiesin easternChina. Considering the uncertainty of delineating of aerosols climate influenceby climate model, in this work, we focus our attention on discussing the combined climate effects of vegetation and CO2 level changes by introducing modern land use and land cover (LULC, includingagricultural vegetation, Liu, 1997) and Pre-Industrial LULC (Yu, et al., 2001), and the corresponding CO2 concentrations respectively.
2. Model and experiment setup
The Version 3 of PSU/NCAR fifth-generation Mesoscale Model (MM5v3), which is non-hydrostatic, including a parameterization of detailed land surface processes, has been used by many researchers to simulate regional climate changes. These works indicated that, the MM5v3 could be well used as a climate model for regional climate simulation after a land surface processes model has been introduced (Liang et al., 2001; Chen et al., 2001; Tang et al., 2003; Solman et al., 2007; Trusilova et al., 2008). Meanwhile, because of its non-hydrostatic frame, the MM5v3 can carry out simulations with higher resolution which is convenient for delineating the processes of small area hydrology recycling.
In this work, the simulation domain mainly covers east areas of China (including the Sichuan basin, SouthwestChina, South China, North China and the parts of theInner Mongolia). The simulated results of NCAR CCM3 are taken as the initial fields and lateral boundary forcing fields, which are used to drive the regional model MM5v3+ Noah LSM.For better understandingthe observed climate changes, we conducted3 experiments, i.e., control run (EPD, adopting the modern LULC distribution, while the CO2 level is 368ppm of 2000 yr); present climate test run (EPDtest, using the modern LULC distribution, while the CO2 level is 280ppm of1870 yr); Pre-industrial sensitive experiment (EBI, adopting the Pre-industrial LULC with the CO2 level 280ppm). In these experiments, the same LULC distribution and CO2 level are adopted for global simulations and corresponding regional one.
In this study, the regional model MM5v3 takes a horizontal grid space of 90km, nested with a 30km grid space sub area (Fig 1). Effective grid numbers are 34*31 and 25*19 in90km and 30km resolution simulationsrespectively, after five-grid-buffer-zone are removed in all these two areas. The atmospheric radiation parameterization of the model is based on CCM2 radiation scheme,planetary boundary layer processes is parameterized using MRF boundary layer scheme, and cumulus parameterization using Kain Fritsch (KF) cumulus scheme. The land surface scheme which incorporates with MM5v3 is Noah LSM (a developed OSU LSM Version, Tang et al., 2003).
The results from EPD show that the combination of these parameterization schemes is appropriate to be employed in Eastern Asia climate simulations, which are consistent with the work of Tang et al. (2003). The modern snow cover isused in each experiment. The duration of each experiment was 80 months started from January 15, of which the first 12 monthsare discarded (except in Fig2.a) to account for model spin-up time. The modern LULC (1990’s) and Pre-industrial LULC(1870’s, Yu, et al., 2001) are interpolated from the same resolution datainto the model grid. The LULC distribution maps at EPD and EBI are plotted in Fig. 1. The comparison of the LULC distribution between modern and the Pre-industrial in the model domain show marked differences, e.g., Eastern China and northern part of Korea Peninsula farming vegetation have been notably increased, while the desert and bareland areas further extended in Shanxi, inner Mongolia and Gansu province.
Fig.1. Distribution of LULC types at present day (a) and Pre-Industrial Revolution (b). Index represent: 1 - broadleaved evergreen forest; 2 - broadleaved deciduous forest; 3 - mixed forest.; 7 - semi-desert; 11 - bare soil/desert; 12 – crop; 15 – water. In Fig. 1a, the squares denote the southwest of China (SW), Sichuan-Shanxi area (SCSX), lower reaches of theYellow River and Huai River valleys (YHV) and eastern China (EC) respectively. In Fig. 1b, the square denotes the nesting area (D2).
The areas for analysis of water balance are marked with rectangle frames in Fig. 1a. The selected domain is based on climatic regimes and covers of major valleys and terrains in China. The territory of central and eastern China is divided into four areas, i.e., the Sichuan-Shanxi section (SCSX) including the main part of the Sichuan province and its north areas, the southwestern China section (SW), the eastern China section (EC) including the main part of the Yangtze-Huaihe valley and the part of southern China, lower reaches of the Yellow River and Huai River valley (YHV).
3． Model results
Figure 2a depicts the temporal variations of Yangtze-Huai River valley averaged precipitation, 2m air temperature in EPD control run, showing that the simulated temperature have evident feature of seasonal variations, and have shorter spin-up time, each year have resemblingamplitude of temperature variations. The precipitation of Yangtze-Huai valleys also exist evident seasonal variations, each year have a precipitation peak in summer, along with the advance and retreat ofsubtropical high in this area, which according with the observed data (Xie et al., 2007). Figure 2b presents a time series of 5-years mean (1995yr - 1999 yr) air temperature of Yangtze-Huai River valley (upper), Sichuan area (middle) and southwest China (lower), showing that, the errors between simulation and observation (Xu et al., 2009) less than 3°C in most seasons. And lots of sensitivity experiments indicated that this kind of deviation is a systematic error and does not seriously affect the relative comparison among simulations. From the EPD control run, we find that the MM5v3 has higher ability to imitate seasonal evolution of precipitation and the onset/march of East Asian summer monsoon.In addition, the model can reproduce reasonable climate response to the changes of the outer forcing too, suggesting that the model is capable of simulating the characters of the East Asia monsoon and the climate changes affected byhuman activities.
Fig.2. Simulated and observed temperature(units: oC), and precipitation (units: mm d-1) changes. a. Yangtze River – Huai River valley averaged 2m air temperature (squares line, upper), and precipitation (circles line, lower); b. Simulated (crosses line) and observed (dots line) seasonal variations of air temperature in Yangtze River – Huai River valley (upper), Sichuan area (middle) and southwest China (lower); c. EPDtest-EBI annual mean surface temperature difference; d. EPD-EPDtest annual mean surface temperature difference; e. EPD-EBI annual mean surface temperature difference; f. same as e. but for summer. Shadings indicate significance at the 95% level. Abscissa denotesyear of simulation in Fig. a, month in Fig. b.
3.1 Temperature
Fig. 2c gives the air temperature changes of EPDtest-EBI, which mainly represents the effect of LULC changes(expanding of farming vegetation and desertification)on temperature, from it we can see that changes of LULCduring the last 130 yrsinduce temperature rise in mostly areas of east of 110oE, while temperature decline in west of this area, especially in mid- and high latitude. The pattern of temperature changes is not well according with the reconstructed temperature data. The temperature difference of EPD-EPDtest shows (Fig. 2d) that, elevated CO2 concentrations would cause temperature risenearly in all regions of East Asia, especially in south of China. This result also present a discrepancy to the fact that the north of China revealed more remarkable temperature rise than in south in recent 100 years (Qin et al., 2005).
By Analyzing the differences of annual mean land surface air temperature between EPD and EBI (Fig. 2e), we can see that changes of LULCdue to industrial process and increasedCO2 concentrations would cause remarkable temperature rise in many regions of East Asia, especially in mid- and high latitude, such as Inner Mongolia, Eastern China and north of Northern China,with maximum over 4oC,while in north of Sichuan province, Gansu province and Guangxi Zhuang Autonomous Region a decreased temperature occurred. Observational data also show that China temperature has increased more outstanding in high latitude than in lower latitude in recent 100 yrs, temperature rise in the North is earlier than in the South (Wang et al., 2002). The most warming areas are in North, Northwest and Northeast China. In north of Northern China and Inner Mongolia, temperature has increased more than 4oC, while Sichuan and Guizhou province show a cold trend (Qin et al., 2005). The temperature simulation are basically consistent with the corresponding observation data,indicating that MM5v3 + Noah LSM can properly simulate the Eastern Asia climate changes which are strongly affected by human activities after integrating the variations of LULC and CO2 concentrations.
From the simulation results of seasonal variations, we can find that summer temperaturepresents a most outstanding rising among all seasons (Fig. 2f). It is not well consist with the observation fact that the most distinguished temperature increase occurred in winter (Dec-Jan-Feb) in recent 100 yrs (Qin et al., 2005). The discrepancy may be due to exaggerated vegetation degradation of EBI experiment in some senses and sensitive effect of land-use changeson RCM. This result agrees with our former study of the effects of vegetation changes on East Asiaclimate changes (Zheng et. al, 2002b). Comparing to the surface air temperature (about 2m height), variations of the land surface temperature is more intensity. Land surface temperature increases in mid- and high latitude might over 10oC in summer(Jun-Jul-Aug), figure not shown(for brevity). With the geopotential height increase, amplitude of temperature changes is gradually reduced. Mid- and low level (under 700hPa) atmosphere temperature has a similar tendency to the land surface, i.e., it also presents a character of outstanding increase in summer, and the changes are more prominent in high latitude than in lower latitude. But the air temperature changes in mid- and upper level have a rather distinct difference fromthat of the low level, its temperature rise in winter and declines in summer. And the annual mean temperature appears to drop (figure omitted) in many areas (especially in mid- and high latitudes). This might be related to the weakened energy transportation from lower layer atmosphere to mid- and upperlevel atmosphere due to vegetation degradation.
3.2 Precipitation and P-E
Precipitation is one of the most important climate elements forEast Asia monsoon system. It reflects the integrated character of general circulation, temperature and atmosphere humidity, etc. Fig 3a depicts the differences of annual mean rainfall between EPD and EBI. It shows that changes of LULC and CO2 level in recent 130 years havebrought a drought trend in some areas of East Asia; evidence in Sichuan basin, Yunnan province, Huai River valley and Shandong byland are most notable, with a central precipitation decrease over 0.6 mm/d, but along the middle reaches of the Yangtze River,the middle reaches of the Yellow River and some areas of south Chinapresent an increasing tendency in precipitation. The corresponding historical rainfall data also shows that, in middle of Northeast China, the middle and lower reach of Yangtze River valley and Southeast coastal areas, the precipitation increase in recent 100 years, while in most parts of Northern China, north of Eastern China and east of Northeast China, rainfall has a marked decrease. In western China, it has more raised precipitation, especially in Xinjiang Uyghur Autonomous Region. But the precipitation in Southwest Chinahas a decreasing tendency (Qin et al., 2005). According to the historical rainfall data, the simulated precipitation variations basically reflect the rainfall change trend ineast areas of Chinaduring the recent 100 years.