【作者】 张蒙蒙；

【导师】 江志红；

【作者基本信息】 南京信息工程大学 ， 气候系统与全球变化， 2012， 硕士

【摘要】 本文利用国家气象信息中心研制的地面和卫星融合的降水资料,通过CMORPH卫星降水资料、融合降水资料与地面观测降水资料的对比,评估融合降水资料的质量,分析融合降水资料精度的时空分布特征,研究影响融合降水资料精度的主要因素,评估融合资料在各区域的降水日变化特征和对持续性降水过程的监测能力,得到以下主要结论：一、对融合降水误差的时空分布特征分析表明；融合降水资料有效利用地面观测降水和卫星降水各自的优势,在降水量值和空间分布上更为合理,误差随时问变化平缓,平均偏差在0值附近摆动,空间相关系数大多稳定在0.9以上；空间分布上,融合降水整体上低估了全国各区域的降水,但与卫星降水相比,融合降水的误差在全国范围内普遍减小,区域性差异减弱,空间一致性提高；除青藏地区外,融合降水与地面观测降水在各区域有较高的空间一致性,全国平均空间相关系数高达0.983；江淮区域的偏差改善尤为明显；与卫星资料相比,融合降水在各月份均以较小偏差占主要比例；融合降水还高估了降水日数,对2mm／d以下的小降水量存在高估,低估2mm／d以上的降水量；与卫星降水相比,融合降水对各强度降水的偏差显著减小,10mm／d以上强度的降水偏差改善明显,尤其对30mm／d以上的强降水,与卫星降水相比误差减小了4-5倍；且融合降水在不同累积时问下的误差与卫星降水相比均大幅度减小,均方根误差减小了3倍,空间相关系数提高了4倍左右。且累积时次越长,误差改善越明显。二、分析站点疏密和地形变化对融合降水精度的影响表明：稀疏区域和密集区域的融合产品质量都与站网密度有关,站网越密集融合效果越好。稀疏区和密集区域的卫星资料在融合产品中的重要性不一样：卫星资料在密集区域的融合作用不大；而在西部没有站点的地方,融合资料主要取决于卫星资料,网格内有观测站时,稀疏区域的融合资料主要取决于地面观测站资料,且网格内站点数越多,融合与地面观测的误差越小。融合效果还与地形高度变化有关,融合降水的量值在地形变化平缓区域与地面观测降水更贴近,空间一致性较好；地形变化复杂区域,低洼处的量值比山峰处的值更接近地面观测降水。CMORPH卫星降水的量值与地形关系不大,但是其空间分布特征在低洼处比山峰处更接近地面观测降水。三、对比融合降水与观测资料的日变化特征,发现：除青藏高原地区外,融合降水能很好的反映平均降水量在各区域的日变化情况；而CMORPH卫星降水未能体现江淮、华南的双峰特征,在其余地方的降水峰值或提前或推后。四、持续性降水过程个例研究表明,融合降水对江淮和华南地区的两次持续性降水过程的误差值分别为-0.016、-0.02mm/h,与CMORPH卫星降水相比偏差显著减小,对实际降水量的再现能力强,能准确反映持续性降水过程。对于超强台风“凡亚比”登陆过程,融合降水能较好再现登陆前、中、后不同时间的降水精细结构,对于定量监测降水过程具有优势。更多还原

【Abstract】 Based on the merged precipitation observed by Automatic Meteorological Stations(AWS) and retrieved by the satellite of CMORPH, the merged precipitation product at hourly/0.1°at/lon temporal-spatial resolution is developed by National Meteorological Information Center. In this paper, the quality of the product is assessed from the points of bias temporal-spatial distribution, the diurnal cycle of precipitation over each area, monitoring capacity of the persistent precipitation process. The main factors affect the accuracy of the merged precipitation are also discussed here. The main results of this research can be summarized in the following items:1) The merged product is effectively used the advantages of the AWS observations and satellite product of CMORPH and more reasonable both at the precipitation quantity and spatial distribution. Bias of the merged product is decreased remarkably, the regional bias and relative error is-0.28and-4.4%respectively. The extent of change over temporal-spatial distribution is weakened meanwhile the spatial correlation coefficient reaches0.983.; Bias of merged product are smaller than CMORPH in the same precipitation order, and the quality is improved further with the precipitation order increases especially the order reaches30mm/d. Along with the increase of total time, the result of merged product gets more significant improvement.2) Station intensity plays an important role both in dense-stationed area and sparse-stationed area, the more gauges, the smaller bias between the merged product and the AWS observations. And terrain height affects the accuracy of the merged product, more gentle the terrain height changes, the smaller bias between the merged product and the AWS observations. While the concern between satellite product of CMORPH and topographic variation is not big.3) Expect the Tibetan Plateau, the merged product show the diurnal cycles of precipitation over other regions very close to the AWS observations. While the satellite product of CMORPH failed to capture the major morning peak over Yangtze Rivers and South China, and show the precipitation peak earlier or later over other regions.4) The merged product can capture the precipitation process very well, the mean bias of two rainfall process over Yangtze River and South China is only-0.016、-0-02mm/h, respectively. The merged product have a definite advantage in the quantitatively rainfall monitoring, can show the rainfall structure of super typhoon "FANAPI" of different landing process.更多还原