| 利用集合均方根卡尔曼滤波反演重庆地区SO2源排放 |
| |
| 引用本文： | 吴钲,谢旻,高阳华,芦华,赵磊,高松.利用集合均方根卡尔曼滤波反演重庆地区SO2源排放[J].环境科学研究,2018,31(1):25-33. |
| |
| 作者姓名： | 吴钲 谢旻 高阳华 芦华 赵磊 高松 |
| |
| 作者单位： | 1.重庆市气象科学研究所, 重庆 401147 |
| |
| 基金项目： | 国家科技支撑计划项目（2014BAC16B06）；重庆市气象局开放式研究基金项目（KFJJ-201607） |
| |
| 摘 要： | 传统的"自下而上"清单方法估算的排放清单，其数据的准确性和时效性存在较大局限.基于集合均方根卡尔曼滤波的源清单反演方法，结合WRF-CMAQ（天气研究和预报模式-公共多尺度空气质量模型）被用于对以清华大学编制的2010年MEIC（中国多尺度排放清单模型）排放清单为基础制作的重庆地区SO2排放源进行反演试验以解决准确性和时效性问题，试验时间段为2014年10月15-31日，重庆主城17个环境空气质量国控监测点ρ（SO2）小时观测资料用于反演及检验.结果表明:该方法能够反演重庆地区SO2源排放量，随着反演次数增加，基于反演排放源预报的ρ（SO2）预报误差持续减小，反演4次后预报误差达到比较低的稳定的水平，其均方根误差均低于20 μg/m3. 5次反演后SO2源排放量用于2014年10月24-29日每天起始预报，其预报的站点、时间平均的均方根误差从100~400 μg/m3降至30 μg/m3以下.反演中应用局地化尺度减少集合取样误差影响，54与81 km两个局地化尺度反演结果对预报改善效果相当，表明主要影响重庆主城ρ（SO2）的源排放位于主城及周边地区，也说明内源排放对重庆主城ρ（SO2）起主要影响.反演后面源排放量主城区降幅约为30 kg/（d·km2），周边地区减少10~20 kg/（d·km2），主城区部分SO2点源排放量降幅约为25 kg/（d·km2），说明2010年MEIC排放清单高估了试验时段重庆地区的SO2排放.
|
| 关 键 词： | SO2 WRF-CMAQ 源排放反演 集合均方根卡尔曼滤波 |
| 收稿时间： | 2017/5/16 0:00:00 |
| 修稿时间： | 2017/9/19 0:00:00 |
Inversion of SO2 Emissions over Chongqing with Ensemble Square Root Kalman Filter |
| |
| WU Zheng,XIE Min,GAO Yanghu,LU Hu,ZHAO Lei and GAO Song.Inversion of SO2 Emissions over Chongqing with Ensemble Square Root Kalman Filter[J].Research of Environmental Sciences,2018,31(1):25-33. |
| |
| Authors: | WU Zheng XIE Min GAO Yanghu LU Hu ZHAO Lei and GAO Song |
| |
| Affiliation: | 1.Chongqing Institute of Meteorological Sciences, Chongqing 401147, China2.Nanjing University, Nanjing 210093, China |
| |
| Abstract: | Emissions inventories based on the 'bottom-up' inventory method have many problems with respect to accuracy and time. To solve this problem, an inversion of the SO2 emission over Chongqing from October 15th to 31st, 2014 was performed with WRF-CMAQ coupled with an ensemble square root Kalman filter. The initial SO2 emissions estimates come from the 2010 Multi-resolution Emission Inventory for China developed by Tsinghua University. Hourly observations of SO2 concentrations at 17 state air quality monitoring stations over the downtown area of Chongqing were used for inversion and verification. The results show that the ensemble square root Kalman filter is suitable for inverting SO2 emissions over Chongqing. The forecast error decreased gradually as the number of inversion emissions increased; additionally, the root mean square errors of the forecast concentration of SO2 were all less than 20 μg/m3 after four inversions. Localization was employed to account for ensemble sampling error. The experiments using a localization distance of 54 and 81 km had similar results, demonstrating that the concentrations of SO2 in Chongqing are primarily affected by the SO2 emission inventories surrounding the area and that the inner sources are primarily responsible for the concentration of SO2 in the downtown area of Chongqing. The average root mean square error of several station forecasts, initialized at each day from October 24th to 29th, 2014, decreased from 100-400 μg/m3 to concentrations less than 30 μg/m3. The estimated SO2 emissions of the area decreased by approximately 30 kg/(d·km2) and 10-20 kg/(d·km2) in the downtown and surrounding areas of Chongqing, respectively; meanwhile, point emissions decreased by approximately 25 kg/(d·km2) in part of the downtown area. The regional SO2 emissions inventories from 2010 MEIC were overestimated over Chongqing during the experiment. |
| |
| Keywords: | SO2 WRF-CMAQ inversion estimation ensemble square root Kalman filter |
| 本文献已被 CNKI 等数据库收录！ |
| 点击此处可从《环境科学研究》浏览原始摘要信息 |
|
点击此处可从《环境科学研究》下载全文 |
|
