南京信息工程大学Ping Zhao的研究团队报道了最大熵产生法在WRF诺亚陆地地表模型中的应用。2023年2月21日，国际知名学术期刊《JGR大气》发表了这一成果。

最大熵产(MEP)模型是一种计算地表热通量的新方法，它与在天气研究中的诺亚陆地地表模型(LSM)和天气研究与预报(WRF)模型中相结合。以2015年6月1日至8月31日青藏高原9个密集观测站和453个常规气象业务观测站的地表热通量和包括气温、相对湿度、土壤温度、降水等气象变量为主题，对耦合模型进行了评价。

研究结果表明，最大熵产方法改善了天气研究与预报模型地表能量平衡的非封闭性(能量余量从24.3 W m^-2增加到1.9 W m^-2)，减少了青藏高原复杂地形对地表感热和潜热通量的高估和对地表地热流的低估(地表感热和潜热通量的偏差分别减少了50.6%和117.1%)，尤其是在白天。改进后的地表感热和潜热通量使青藏高原的冷偏和湿偏分别降低了29.4%和51.7%。最大熵产模型还将高估的土壤温度降低了63%。此外，偏高的日降水量比青藏高原降低了3%。最大熵产的方法的成功应用证明了其相对于整体传输方法的优势，为减少复杂地形下数值预报模型对地表热通量的高估和干湿偏差提供了一种新的方法。

在数值天气和气候模式中，利用复杂地形、高植被区和早晚过渡时期的整体传输模式计算地表感热(SH)和潜热(LE)通量，仍然是一个具有挑战性的问题。

附：英文原文

Title: An application of the maximum entropy production method in the WRF Noah land surface model

Author: Chunhui Jia, Ping Zhao, Jingfeng Wang, Yi Deng, Na Li, Yingchun Wang, Shiguang Miao

Issue&Volume: 2023-02-21

Abstract: The calculation of surface sensible heat (SH) and latent heat (LE) fluxes using the bulk transfer models for complex terrains, tall vegetation regions, and morning and evening transition periods remains a challenging problem in numerical weather and climate models. The maximum entropy production (MEP) model, a new method of calculating surface heat fluxes, is coupled with the Noah Land Surface Model (LSM) in the Weather Research and Forecasting (WRF) model. Surface heat fluxes and meteorological variables including air temperature, relative humidity, soil temperature, and precipitation data at nine intensive observation stations and 453 routine meteorological operational observation stations in the Tibetan Plateau (TP) from 1 June to 31 August 2015 are used to evaluate the coupled model. The results show that the MEP method improves the nonclosure of the surface energy balance in the WRF (with an energy residual from 24.3 W m^-2 to 1.9 W m^-2), reducing the overestimations of SH and LE and the underestimation of the surface ground heat flux over the complex terrain of TP (with the bias decreases of 50.6% and 117.1% for SH and LE in turn), especially during the daytime. The improved SH and LE reduce the cold and wet biases in the TP by 29.4% and 51.7%, respectively. The MEP model also reduces the overestimated soil temperature by 63%. Moreover, the overestimated daily precipitation is reduced by 3% over the TP. The successful application of the MEP method demonstrates its advantage over the bulk transfer method, providing a new approach for reducing the overestimation of surface heat fluxes and wet and cold biases in numerical forecast models in complex terrains.

DOI: 10.1029/2022JD037867

Source: https://agupubs.onlinelibrary.wiley.com/doi/10.1029/2022JD037867