基于陆面模式的干旱监测技术及其在我国的应用效果研究

发布时间：2018-07-10 05:30

  本文选题：干旱监测 + 陆面模式　； 参考：《兰州大学》2016年博士论文

【摘要】：干旱是一种频繁发生的自然灾害,长时期、大范围观测资料尤其是土壤湿度资料的缺乏制约着人们对干旱问题的认识和监测与评估。近年来,陆面模式结果作为一种资料替代正受到越来越多的重视和应用。而在我国,能够利用陆面模式进行实时监测干旱的工作还较为缺乏。因此,本文利用CABLE(Community Atmosphere Biosphere Land Exchange)陆面模式,在对模式模拟效果检验基础上,建立基于CABLE模式的干旱监测技术和方法。通过从干旱定义和物理过程复杂性方面与SPI、SPEI和PDSI指数进行对比分析并验证模式监测效果。并分区域、分季节,从干旱演变趋势、受旱面积、累积发生月数、持续时间和影响机理方面,探讨我国过去60年干旱演变特征。主要结论包括:(1)模式模拟能力评估。CABLE模拟的蒸散发和总径流量与GSWP-2多模式结果和全球主要河流观测值一致性高。在热带地区,CABLE与GSWP-2对蒸散发的低估由降水强迫资料误差引起。与VIC模拟结果和通榆站观测资料比较显示,CABLE能够再现我国地表径流和蒸发的基本特征。对我国中东部土壤(0~173.8 cm)湿度缓变趋势模拟效果好。在西北地区,春季积雪融化对模拟土壤湿度影响明显。(2)准实时强迫资料构建和基于CABLE模式的干旱监测技术开发。应用原有1948-2000年NCC(NCEP Corrected by CRU)强迫资料构建方法,利用CMAP周平均降水强迫资料替代原CRU降水资料,将NCC资料从2001年延伸到准实时(滞后7 d左右)状态。延伸后与延伸前及观测降水资料比较,资料偏差、均方根误差和相关系数均表明延伸后资料的可靠性。华南地区误差较大由CMAP资料造成。并综合运用Shell语言、NCO和GrADS软件,完成基于CABLE陆面模式的干旱监测技术开发,并实现从强迫资料生成到监测产品绘图各流程自动化。(3)对我国近年来重大干旱事件监测效果检验。对河南省2001-2002年典型干旱事件分析显示,模式能够有效监测河南省干旱事件的发生和发展过程,与各区域站点对比显示出模式对干旱起止时间和等级的监测能力。对2009-2011年西南(贵州)分析显示,模式监测干旱发生时间滞后于K和CI指数,干旱持续时间较两个指数短。西南地区土壤雨季得到充分补充能够延滞干季土壤水分流失速度和K与CI指数侧重监测气象干旱是造成差异原因。(4)与标准化降水指数(SPI),标准降水蒸发指数(SPEI)和帕尔默旱度指数(PDSI)的对比分析,揭示出不同干旱指数定义和物理过程复杂性对干旱监测效果的影响。对由同一套NCC强迫资料得到的干旱指数对比显示,我国大部分地区春、秋季SPI和SPEI空间分布相似,表明干旱主要受降水影响。由于西北春、秋季温度对蒸发影响强于降水,所以PDSI和CABLE对蒸发影响因素的更全面考虑使春、秋季两者分布特点相似。夏季,PDSI和CABLE在南方的分布差异由该季节地表能量对蒸发起主导作用、蒸发计算方法不同、植被覆盖度和土壤蒸散和冠层传输差异共同造成。通过对1997/1998年和2009/2010年两次干旱个例的分析表明,模式与观测土壤湿度结果一致、与PDSI监测效果接近。CABLE还能够反映出不同土壤层的变化和深层土壤需要更多降水和时间补充植被根区水分流失这一现象。(5)揭示了过去60年的干旱演变特征。1990s中期开始,东北、华北和西南显著变干,干旱发生次数增多;西北和西藏2000年后略微增湿,干旱发生次数减少。1970s中期和1990年左右是全国受旱面积较少时期。2000年后,受旱面积显著增长。1954-2013年,我国轻-特旱年均受旱面积分别占占国土面积15.01%、10.02%、5.02%和5.07%,年均受旱面积为35.12%。东北、西南和华中受旱面积2000年之后显著增加,华北1995年后表现为增加趋势。西北和西藏地区无明显变化,华东略有增加。前(1954-1983年)后(1984-2013年)两个时期,干旱累积发生月数分析显示,干旱增幅度最大在100°E以东北方地区。区域性重旱发生次数由多到少依次为:东北、华中、华北、华东、西南和西藏。且东北、西南和华东发生在2000年之后。(6)揭示了影响我国干旱演变的主要因素。春季,各区域土壤湿度与降水均为正相关,相关系数为0.44~0.67。东北和西藏由于温度较低,引起模式中土壤结冰、土壤孔隙度减小,降水下渗减少,土壤水分补充减少,加之冻融过程影响,造成降水与土壤湿度变化相关性低。短波辐射与土壤湿度在大部分地区呈负相关。东北、西北和西藏春季气温低并常有积雪,地表反照率高,接收能量减少,造成辐射与土壤湿度相关性不高。秋季与春季类似,但受“华西秋雨”和东北秋季降水偏多影响,西藏和东北土壤湿度与降水相关性高于春季。且秋季由于生长旺盛的植被能够阻止土壤水分流失,造成大部分地区土壤湿度与短波辐射相关性不如春季显著。夏季,南方土壤水分呈饱和状态与多云天气影响太阳辐射能量传输和吸收是造成南方土壤湿度与降水和辐射相关性低于北方的主要原因。
[Abstract]:Drought is a frequent natural disaster. For a long period of time, the lack of extensive observation data, especially the lack of soil moisture data, restricts people's understanding and monitoring and evaluation of drought. In recent years, land surface model results are becoming more and more reconsidered and applied as a kind of data replacement. In China, land surface model can be used. There is still a lack of real-time monitoring of drought. Therefore, this paper uses the land surface model of CABLE (Community Atmosphere Biosphere Land Exchange) to establish a drought monitoring technique and method based on the model simulation effect on the basis of the model simulation effect. Through the drought determination and the physical process complexity, SPI, SPEI and PDSI refers to SPI. A comparative analysis was carried out and the effect of model monitoring was verified. The characteristics of drought evolution in China in the past 60 years were discussed in the areas of drought evolution, drought area, cumulative occurrence months, duration and influence mechanism. The main conclusions were as follows: (1) model simulation ability evaluation of.CABLE simulated evapotranspiration and total runoff and GSWP-2 The model results are consistent with the global major river observation values. In tropical areas, the underestimation of Evapotranspiration by CABLE and GSWP-2 is caused by the error of precipitation forcing data. Compared with the VIC simulation results and Tongyu station observation data, CABLE can reproduce the basic characteristics of surface runoff and evaporation in China. The humidity of 0~173.8 cm in the Middle East of China The effect of slow variation trend is good. In Northwest China, the influence of spring snow melting on simulated soil moisture is obvious. (2) quasi real time forcing data construction and the development of drought monitoring technology based on CABLE model. The original 1948-2000 year NCC (NCEP Corrected by CRU) forced data construction method is used to replace the original CRU by means of CMAP weekly mean precipitation forced data. Precipitation data, extending the NCC data from 2001 to quasi real time (lagging behind 7 d). After extension, data deviation, root mean square error and correlation coefficient all indicate the reliability of the data after extension and observed precipitation data. The error of Southern China area is larger by CMAP data. Shell language, NCO and GrADS software are used in a comprehensive way. The drought monitoring technology based on the CABLE land surface model has been developed and realized from the forced data to the monitoring product drawing automation. (3) the monitoring results of the major drought events in China in recent years are tested. The analysis of the typical drought events in 2001-2002 years in Henan shows that the model can effectively monitor the occurrence and development of drought events in Henan province. The analysis of 2009-2011 year Southwest (Guizhou) shows that the occurrence time of the model monitoring drought is lagging behind K and CI index, and the duration of drought is shorter than that of the two index. The soil moisture in the southwest of the dry season can delay the dry season soil moisture. Loss rate and K and CI index focus on monitoring meteorological drought. (4) compared with standardized precipitation index (SPI), standard precipitation evaporation index (SPEI) and Palmer drought index (PDSI), the effects of different drought index definitions and physical Guo Chengfu miscellaneous properties on drought monitoring are revealed. The comparison of the drought index obtained by the data shows that the spatial distribution of SPI and SPEI in autumn is similar in most of China's spring and autumn, which indicates that drought is mainly affected by precipitation. Due to the northwest spring, autumn temperature has a stronger influence on evaporation than precipitation, so PDSI and CABLE have more comprehensive consideration of the influence factors of evaporation in spring and autumn. In summer, PDSI and CAB The distribution difference of LE in the south is dominated by the surface energy of this season, and the evaporation calculation method is different. The vegetation coverage and the difference of soil evapotranspiration and canopy transmission are common. The analysis of the two drought cases in 1997/1998 and 2009/2010 shows that the model is consistent with the observation of soil moisture and is close to the effect of PDSI monitoring. .CABLE can also reflect the changes in different soil layers and the need for more precipitation and time to supplement the water loss of vegetation roots. (5) it reveals that the characteristics of drought evolution in the past 60 years have begun in the middle of the middle period of.1990s, the northeast, North and southwest significantly dry, the number of droughts increased, and the northwest and Tibet slightly humidified after 2000. The number of droughts decreased in the middle of.1970s and 1990 was.2000 years in the less drought area in China. The drought area increased significantly for.1954-2013 years. The annual drought area of light and special drought in China accounted for 15.01%, 10.02%, 5.02% and 5.07% respectively, and the average annual drought area was 35.12%. The northeast, southwest and central China were significantly affected by the drought area after 2000. In North China, there was an increasing trend after 1995. There was no obvious change in the northwest and Tibet regions, and the East China increased slightly. After two periods (1954-1983 years) (1984-2013 years), the drought accumulation month analysis showed that the maximum increase of drought was in the north of 100 E. The number of severe drought occurred in the northeast and central China, in turn. North China, East China, southwest and Tibet. And northeast, southwest and East China after 2000. (6) revealed the main factors affecting the evolution of drought in China. In spring, the soil moisture and precipitation in each region are positively correlated, the correlation coefficient is 0.44~0.67. northeast and Tibet because of the low temperature, causing the soil ice in the model, the decrease of soil porosity and precipitation. The decrease of infiltration, the decrease of soil moisture supplement and the influence of the freezing thawing process, resulting in low correlation between precipitation and soil moisture change. The short wave radiation and soil moisture are negatively correlated in most areas. In the northeast, northwest and Tibet, the temperature is low and often has snow, the surface albedo is high and the receiving energy is reduced, and the correlation between radiation and soil moisture is not related. Autumn is similar to spring, but influenced by "West autumn rain" and precipitation in Northeast autumn, the correlation of soil moisture and precipitation in Tibet and northeast is higher than that in spring. And the growing vegetation in autumn can prevent soil moisture loss, and the correlation between soil moisture and short wave radiation in most areas is not as significant as that in spring. The saturated state of soil moisture and the influence of multi cloud weather on the transmission and absorption of solar radiation energy are the main reasons for the lower correlation between soil moisture and precipitation and radiation in the south.
【学位授予单位】：兰州大学
【学位级别】：博士
【学位授予年份】：2016
【分类号】：P426.616;P412
，

本文编号：2112245