Research on improvement of temperature forecasts of the regional numerical prediction system using CLDAS land data

doi:10.11755/j.issn.1006-7639(2024)-02-0238

Abstract

Abstract:

In order to improve the performance of the regional mesoscale numerical prediction system on temperature forecasts, the CLDAS (CMA Land Data Assimilation System) land data in Zhejiang Province is evaluated for its precision using observed soil temperature data as well as soil moisture data, and then it is applied in Zhejiang numerical prediction system. The results are shown as follows: The CLDAS soil temperature and soil moisture products have smaller root mean squared errors and higher correlation coefficients compared to the observations than those products of the GFS (Global Forecast System) analysis field, and they have good applicability in Zhejiang Province. The case study indicates that 2 m temperature predicted by the regional numerical model is sensitive to the change of land data. It exerts a sustainable influence on the model forecasts at a later stage through blending the CLDAS real-time surface temperature, soil temperature and soil moisture analysis products on the initial field. More specifically, the temperature changes are directly affected by the surface sensible heat flux and latent heat flux. Besides, the root mean squared error of the model prediction using the CLDAS data is lowered by 13.6% in comparison to that of the model prediction using the GFS analysis field as the initial land surface field. The phased application results in July 2021 show the model initial field blending the CLDAS data can effectively improve the forecast accuracy of the 2 m temperature in Zhejiang Province. Moreover, the results present a more favorable improvement at night than in the daytime, and more effective forecasts can be presented under dry-hot weather condition than tropical cyclones and Mei-Yu fronts periods.The evaluation of high temperature forecasts further indicates that the application of CLDAS land data has a good improvement effect on the predictions of high temperature events in Zhejiang Province, especially in high-temperature areas such as the Jinqu Basin.

Key words: soil temperature, soil moisture, 2 m temperature, high temperature, numerical weather prediction model

摘要：

为提升区域数值预报系统2 m气温预报性能，利用土壤温度和土壤湿度站点观测资料，对中国气象局陆面数据同化系统（CMA Land Data Assimilation System，CLDAS）陆面资料在浙江地区的精度进行评估，并将其融合应用于浙江省数值预报业务系统。结果表明：CLDAS土壤温度、土壤湿度产品相对于美国全球预报系统（Global Forecast System，GFS）分析场，与观测相比具有更小的均方根误差和更高的相关系数，在浙江省有较好的适用性。个例分析表明区域数值模式2 m气温预报对陆面资料变化较敏感，融合CLDAS地表温度、土壤温湿度实时分析产品的初始场，可持续影响到模式预报后期，主要通过地表感热、潜热通量直接影响气温变化。从均方根误差来看，与基于GFS分析场作为陆面初始场的区域模式预报相比，应用了CLDAS陆面资料的模式预报改进了13.6%。2021年7月阶段性应用结果显示，模式初始场融合CLDAS陆面资料后有效提高了浙江省2 m气温预报水平，融合后的预报改进效果夜间较白天明显，且晴热高温天气背景下较梅雨期、台风期改进更佳。高温天气预报评估进一步表明，CLDAS陆面资料的应用对浙江省高温事件预报有较好的改进，尤其对金衢盆地等高温区改进明显。

关键词: 土壤温度, 土壤湿度, 2 m气温, 高温, 数值天气预报模式

CLC Number:

P457.3

QIU Jinjing, CHEN Feng, DONG Meiying, FAN Yuemin, YU Zhenshou. Research on improvement of temperature forecasts of the regional numerical prediction system using CLDAS land data[J]. Journal of Arid Meteorology, 2024, 42(2): 238-250.

邱金晶, 陈锋, 董美莹, 范悦敏, 余贞寿. CLDAS陆面资料对区域数值预报系统气温预报的改进研究[J]. 干旱气象, 2024, 42(2): 238-250.

Figures/Tables 12

Fig.1 Hourly variation of average rainfall （a） and 2 m air temperature （b） in Zhejiang Province in July 2021

Fig.2 The 12-hour variation of averages， absolute errors， root mean square errors and correlation coefficients of 0-10 cm soil temperature （a） and 0-10 cm soil moisture （b） of CLDAS and GFS analysis fields in Zhejiang Province in July 2021

Fig.3 Spatial distributions of time correlation coefficients between CLDAS （a， c）， GFS analysis fields （b， d） and 0-10 cm soil temperature （a， b）， 0-10 cm soil moisture （c， d） in July 2021 in Zhejiang Province

Fig.4 The errors （a）， absolute errors （b）， root mean square errors （c） and correlation coefficients （d） of average 2 m temperature predicted by TCLDAS and TGFS for 1-72 h validity period starting at 20：00 on 13 July 2021 in Zhejiang Province

Fig.5 Spatial distributions of errors of 2 m temperature of 1-72 h average and the 18th hour， the 42th hour and the 66th hour in Zhejiang Province predicted by TCLDAS and TGFS starting at 20：00 on 13 July 2021 （Unit： °C）（Black quadrilateral is the focus area for the following part）

Fig.6 Difference distributions between CLDAS and GFS surface temperature， soil temperature （Unit： ℃） for different layers and soil moisture （Unit： m3·m-3） for different layers at 20：00 on 13 July 2021

Fig.7 Difference distributions of 0-10 cm soil temperature （Unit： ℃）， 0-10 cm soil moisture （Unit： m3·m-3）， sensible heat flux （Unit： W·m-2） and latent heat flux （Unit： W·m-2） for 1-72 h average predicted by TCLDAS and TGFS starting at 20：00 on 13 July 2021

Fig.8 The 1-72 h regional averages of 2 m temperature （a） and their difference （b）， 0-10 cm soil temperature （c） and their difference （d）， 0-10 cm soil moisture （e） and their difference （f）， sensible heat flux （g） and their difference （h）， latent heat flux （i） and their difference （j） predicted by TCLDAS and TGFS starting at 20：00 on 13 July 2021

Tab.1 Evaluation results of 2 m temperature in Zhejiang Province predicted by TCLDAS and TGFS under different weather backgrounds in July 2021

时段	试验名称	误差/℃	绝对误差/℃	均方根误差/℃	相关系数
1—10日	TCLDAS	-0.75	1.82	2.22	0.55
1—10日	TGFS	-1.10	1.92	2.32	0.54
11—19日	TCLDAS	-1.30	1.89	2.31	0.67
11—19日	TGFS	-1.77	2.18	2.61	0.65
20—31日	TCLDAS	-0.23	1.55	1.94	0.63
20—31日	TGFS	-0.60	1.66	2.05	0.62

Fig.9 Comparison of evaluation parameters of 2 m temperature （≥35 ℃） at 14：00 under different lead time for TCLDAS and TGFS in July 2021 （a） hit rate，（b） threat score，（c） frequency bias

Fig.10 Spatial distributions of average 2 m temperature at 14：00 in July 2021 （Unit： ℃）（a） forecasted by TCLDAS 6 h in advance，（b） forecasted by TGFS 6 h in advance，（c） forecasted by TCLDAS 18 h in advance，（d） forecasted by TGFS 18 h in advance，（e） observation

Fig.11 The spatial distributions of averages and differences of threat scores for daily 2 m temperature （≥35 ℃） at 14：00 predicted by TCLDAS and TGFS in July 2021 （a） forecasted by TCLDAS 6 h in advance，（b） forecasted by TGFS 6 h in advance，（c） difference between TCLDAS and TGFS predictions 6 h in advance，（d） forecasted by TCLDAS 18 h in advance，（e） forecasted by TGFS 18 h in advance，（f） difference between TCLDAS and TGFS predictions 18 h in advance

References 41

[1]	陈锋, 冀春晓, 董美莹, 等, 2012a. 雷达径向风速同化对台风麦莎模拟的影响[J]. 气象, 38(10): 1 170-1 181.
[2]	陈锋, 董美莹, 冀春晓, 等, 2012b. WRF模式对浙江2011年夏季降水和温度预报评估及其湿过程敏感性分析[J]. 浙江气象, 33(3): 3-12.
[3]	陈燕丽, 黄思琦, 莫建飞, 等, 2020. 基于CLDAS数据的甘蔗干旱监测评估标准对比——以2011年广西干旱为例[J]. 干旱气象, 38(2): 188-194.
[4]	崔园园, 张强, 李威, 等, 2020. CLDAS融合土壤相对湿度产品适用性评估及在气象干旱监测中的应用[J]. 海洋气象学报, 40(4): 105-113.
[5]	丁金才, 1995. 天气预报评分方法评述[J]. 大气科学学报, 18(1): 143-150.
[6]	董祝雷, 赵艳丽, 冯晓晶, 等, 2023. CLDAS气温和降水产品在内蒙古地区适用性分析[J]. 干旱气象, 41(5): 811-819. DOI
[7]	郭阳, 师春香, 徐宾, 等, 2023. CLDAS陆面融合实况数据对天津雾和霾判识的准确性分析[J]. 干旱气象, 41(4): 657-665. DOI
[8]	韩帅, 师春香, 林泓锦, 等, 2015. CLDAS土壤湿度业务产品的干旱监测应用[J]. 冰川冻土, 37(2): 446-453.
[9]	刘欢欢, 王飞, 张廷龙, 2018. CLDAS和GLDAS土壤湿度资料在黄土高原的适用性评估[J]. 干旱地区农业研究, 36(5): 270-276+294.
[10]	刘佩佩, 宋海清, 鲍炜炜, 等, 2021. CLDAS和GLDAS土壤温度数据在陕西省的适用性评估[J]. 气象科技, 49(4): 604-611.
[11]	马柱国, 魏和林, 符淙斌, 2000. 中国东部区域土壤湿度的变化及其与气候变率的关系[J]. 气象学报, 58(3): 278-287.
[12]	师春香, 姜立鹏, 朱智, 等, 2018. 基于CLDAS2.0驱动数据的中国区域土壤湿度模拟与评估[J]. 江苏农业科学, 46(4): 231-236.
[13]	师春香, 潘旸, 谷军霞, 等, 2019. 多源气象数据融合格点实况产品研制进展[J]. 气象学报, 77(4): 774-783.
[14]	师春香, 谢正辉, 钱辉, 等, 2011. 基于卫星遥感资料的中国区域土壤湿度EnKF数据同化[J]. 中国科学: 地球科学, 41(3): 375-385.
[15]	王莉莉, 龚建东, 2018. 两种OI陆面同化方法在GRAPES_Meso模式中的初步应用试验[J]. 气象, 44(7): 857-868.
[16]	王万秋, 1991. 土壤温湿异常对短期气候影响的数值模拟试验[J]. 大气科学, 15(5): 115-123.
[17]	温晓培, 吴炜, 李昌义, 等, 2022. 土地利用资料的更新对四川盆地高温天气数值模拟的影响[J]. 干旱气象, 40(5): 868-878. DOI
[18]	徐金芳, 邓振镛, 陈敏, 2009. 中国高温热浪危害特征的研究综述[J]. 干旱气象, 27(2): 163-167.
[19]	易翔, 曾新民, 王宁, 等, 2016a. WRF模式中土壤湿度对位势高度模拟影响的敏感性分析[J]. 干旱气象, 34(1): 113-124.
[20]	易翔, 曾新民, 郑益群, 等, 2016b. 高分辨率WRF模式中土壤湿度扰动对短期高温天气模拟影响的个例研究[J]. 大气科学, 40(3): 604-616.
[21]	余贞寿, 冀春晓, 杨程, 等, 2018. 同化风廓线雷达资料对浙江降水预报改进评估[J]. 应用气象学报, 29(1): 97-110.
[22]	曾庆存, 周广庆, 浦一芬, 等, 2008. 地球系统动力学模式及模拟研究[J]. 大气科学, 32(4): 653-690.
[23]	CHEN F, DUDHIA J, 2001. Coupling an advanced land surface-hydrology model with the Penn State-NCAR MM5 modeling system. Part I: Model implementation and sensitivity[J]. Monthly Weather Review, 129(4): 569-585.
[24]	FAN X G, 2009. Impacts of soil heating condition on precipitation simulations in the weather research and forecasting model[J]. Monthly Weather Review, 137(7): 2 263-2 285.
[25]	FERRANTI L, VITERBO P, 2006. The European summer of 2003: Sensitivity to soil water initial conditions[J]. Journal of Climate, 19(15): 3 659-3 680.
[26]	GIARD D, BAZILE E, 2010. Implementation of a new assimilation scheme for soil and surface variables in a global NWP model[J]. Monthly Weather Review, 128(4): 997-1 015.
[27]	HONG S Y, DUDHIA J, CHEN S H, 2004. A revised approach to ice microphysical processes for the bulk parameterization of clouds and precipitation[J]. Monthly Weather Review, 132(1): 103-120.
[28]	HONG S Y, PAN H L, 1996. Nonlocal boundary layer vertical diffusion in a medium-range forecast model[J]. Monthly Weather Review, 124 (10): 2 322-2 339.
[29]	IACONO M J, DELAMERE J S, MLAWER E J, et al, 2008. Radiative forcing by long-lived greenhouse gases: Calculations with the AER radiative transfer models[J]. Journal of Geophysical Research (Atmospheres), 113(D13), D13103. DOI:10.1029/2008JD009944.
[30]	JACOBS C M J, MOORS E J, Ter Maat H W, et al, 2010. Evaluation of European Land Data Assimilation System (ELDAS) products using in situ observations[J]. Tellus A, 60(5): 1 023-1 037.
[31]	JIMéNEZ P A, DUDHIA J, GONZáLEZ-ROUCO J F, et al, 2012. A revised scheme for the WRF surface layer formulation[J]. Monthly Weather Review, 140(3): 898-918.
[32]	LIM Y J, LEE T Y, BYUN K Y, 2008. Korea Land Data Assimilation System (KLDAS) and its application using WRF[C]// In 22nd Conference on Hydrology, 88th American Meteorological Society Annual Meeting, New Orleans, LA.
[33]	MAHANAMA S P P, KOSTER R D, REICHLE R H, 2008. Impact of subsurface temperature variability on surface air temperature variability: An AGCM study[J]. Journal of Hydrometeorology, 9(4): 804-815.
[34]	PAN H L, MAHRT L, 1987. Interaction between soil hydrology and boundary-layer development[J]. Boundary-Layer Meteorology, 38: 185-202.
[35]	PETERS-LIDARD C D, BLACKBURN E, LIANG X, et al, 1998. The effect of soil thermal conductivity parameterization on surface energy fluxes and temperatures[J]. Journal of Atmospheric Sciences, 55(7): 1 209-1 224.
[36]	RODELL M, HOUSER P R, JAMBOR U, et al, 2004. The global land data assimilation system[J]. Bulletin of the American Meteorological Society, 85(3): 381-394.
[37]	WANG X, PARRISH D, KLEIST D, et al, 2013. GSI 3DVar-based ensemble-variational hybrid data assimilation for NCEP Global Forecast System: Single-resolution experiments[J]. Monthly Weather Review, 141(11): 4 098-4 117.
[38]	NIU G Y, YANG Z L, MITCHELL K E, et al, 2011. The community Noah land surface model with multi-parameterization options (Noah-MP):1. Model description and evaluation with local-scale measurements[J]. Journal of Geophysical Research (Atmospheres), 116, D12109. DOI: 10.1029/2010JD015139.
[39]	YANG Z L, NIU G Y, MITCHELL K E, et al, 2011. The community Noah land surface model with multi-parameterization options (Noah-MP):2.Evaluation over global river basins[J]. Journal of Geophysical Research (Atmospheres), 116, D12110. DOI: 10.1029/2010JD015140
[40]	ZENG X M, WANG B, ZHANGY, et al, 2014. Sensitivity of high-temperature weather to initial soil moisture: A case study using the WRF model[J]. Atmospheric Chemistry and Physics, 14: 9 623-9 639.
[41]	ZOU X, QIN Z K, ZHENG Y, 2015. Improved tropical storm forecasts with GOES-13/15 imager radiance assimilation and asymmetric vortex initialization in HWRF[J]. Monthly Weather Review, 143(7): 2 485-2 505.