不同初始场及陆面方案对青藏高原中东部积雪消融过程的模拟研究  

Numerical Assessment of Various Initial Conditions and Land Surface Schemes on the Snow Melting Process over the Central and Eastern Qinghai-Xizang Plateau

作　　者：姜琪 罗斯琼 李明 JIANG Qi;LUO Siqiong;LI Ming(Huangpi Meteorological Bureau of Wuhan,Wuhan 432200,Hubei,China;Key Laboratory of Land Surface Process and Climate Change in Cold and Arid Regions/Northwest Institute of Eco-Environment and Resources,Chinese Academy of Sciences,Lanzhou 730000,Gansu,China;Wuhan Meteorological Bureau of Hubei,Wuhan 430040,Hubei,China)

机构地区：[1]武汉市黄陂区气象局,湖北武汉432200 [2]中国科学院西北生态环境资源研究院/中国科学院寒旱区陆面过程与气候变化重点实验室,甘肃兰州730000 [3]武汉市气象局,湖北武汉430040

出　　处：《高原气象》2022年第2期430-443,共14页Plateau Meteorology

基　　金：国家自然科学基金项目(41975096,U20A2081);中科院寒旱区陆面过程与气候变化重点实验室2019年度开放基金项目(LPCC2019008)。

摘　　要：基于WRF(Weather Research and Forecasting)模式,本研究使用更为准确的气象站点及卫星遥感积雪资料替换初始场中积雪深度、雪水当量等积雪数据,对2014年2月17-27日青藏高原中东部一次积雪消融过程进行模拟研究,评估WRF模式中CLM(Community Land Model)、Noah-LSM(Noah land surface model)和Noah-MP(Noah-Multiparameterization Land Surface Model)3种陆面过程方案对该次积雪消融过程的模拟性能。结果表明:3种陆面过程方案均能较好地再现2 m气温、积雪深度和反照率的日变化趋势,但各试验模拟效果有一定差异。气象站点及卫星遥感积雪资料作为初始场时,CLM陆面过程方案模拟的2 m气温平均误差最小,为0.002℃;Noah-LSM陆面过程方案中2 m气温均方根误差(4.01℃)和平均绝对误差(3.30℃)最小,但昼夜温差较观测显著偏小;同时CLM陆面过程方案模拟的积雪深度均方根误差、平均误差和平均绝对误差均最小,分别为4.70 cm、-1.25 cm和2.75 cm,但该方案对融雪速率有所高估,导致积雪消融时间被低估。Noah-LSM和Noah-MP陆面方案模拟的融雪速率与观测较为一致。3种陆面过程均能较好地再现积雪消融阶段反照率整体呈下降的趋势,但反照率日变化细节均有所欠缺。初始场积雪数据使用ECMWF Reanalysis v5(ERA5)数据集时,CLM陆面过程方案表现最优,再现积雪反照率变化特征,日变化特征与观测较为一致;Noah-LSM和Noah-MP陆面过程方案不能再现反照率日内变化,且低估新雪反照率。CLM陆面过程方案能再现积雪消融过程中净辐射逐渐增大的特征,且峰值与观测相近,同时也能模拟出地表与大气,地表与浅层土壤的能量传输。本研究将为青藏高原积雪模拟陆面方案的选取提供参考依据。In this study,the Weather Research and Forecasting(WRF)model with meteorological observations and satellite-derived snow data incorporated as the initial condition was adopted to simulate a typical snow ablation event that occurred from 17 to 27 February 2014 over the Central and Eastern Qinghai-Xizang Plateau for investigating the effects of various land surface process schemes(Community Land Model,CLM),Noah land surface model(Noah-LSM),and Noah land surface model with multi-parameterization options(Noah-MP)on the snow ablation process. Results showed that WRF model performance varied slightly among different simulation experiments. The diurnal variation of 2-metre temperature,snow depth and albedo were well reproduced by the WRF model with three land surface process schemes. With the integration of an accurate initial snowfield,the CLM land surface process scheme presented the lowest mean bias of 0. 002 ℃ for 2-metre temperature. The WRF simulation with the Noah-LSM scheme had the lowest root mean square error(4. 01 ℃)and lowest mean bias(3. 30 ℃)for 2-metre temperature,but the temperature difference between daytime and nighttime was significantly underestimated. In contrast,the WRF experiment with the CLM scheme achieves excellent performance with the smallest root mean square error,mean error,and absolute mean error of 4. 70 cm,-1. 25 cm,and 2. 75 cm,respectively. However,the CLM experiment overestimated the snowmelt rate,resulting in an underestimation of the snowmelt phase time. Compared with observations,the snowmelt rates simulated by NoahLSM and Noah-MP experiments showed a good agreement with monitoring data. The decreasing trend of albedo during the snow melting stage was well reproduced by these three land surface process schemes,while the diurnal pattern of albedo was not captured by the WRF model attributed to the bias in snow melting speed. With the integration of the snowfield from ECMWF Reanalysis v5(ERA5)dataset as the initial condition,the CLM experiment successfully reproduced the variab

关 键 词：青藏高原 积雪 WRF模式 陆面过程 

分 类 号：P426.635[天文地球—大气科学及气象学]