Application of SSIM in Spatial Testing of High-Resolution Precipitation Numerical Forecasting Products
-
摘要: 引入图像结构相似度指数SSIM应用于高分辨率降水数值预报产品的空间检验，通过对SSIM应用场景和算法的适用性改进，提出了适用于0-1化降水格点场空间检验的SSIM应用方法，并针对SSIM指标不能正确反映降水场中雨带位置变化的缺陷进行了改进，提出了修正的SSIM*指标。应用该新方法选取2020年夏季贵州地区的暴雨个例，对中国气象局/广东省区域数值天气预报重点实验室华南区域中尺度模式输出的高分辨率降水预报产品进行了空间检验，并与采用另外两种检验方法即MODE和邻域法的检验结果进行了比较。结果表明改进的SSIM方法在高分辨率降水数值预报场空间检验中有较好的应用效果，三种方法得到的检验结果在有无预报能力的定性评价上差异不大，而新的SSIM*指标与MODE的MMI指标表现较为接近，且在对个例预报能力的评价上更为合理。该方法以0-1化降水格点场作为空间检验对象，能够得到其它方法所不能得到的降水场整体检验特征和分项检验特征，提供从雨区面积、雨带结构及其平均位置多个方面评估预报性能的参考指标，且计算简单无主观参数设定，得到的检验结果具有唯一确定性，克服了当前MODE方法在应用中的缺点，具有较大的推广应用价值，值得进一步深入研究。Abstract: This study proposes a new scheme for comparing and analyzing 0-1 formatted grid fields using the image structure similarity index method (SSIM). With the original application scenario and algorithms optimized, the SSIM index is improved to reflect the position change of rainbands in precipitation fields. The new method is applied to the spatial verification of high-resolution precipitation numerical forecast products generated by the regional modeling system of South China (CMA-GD), specifically for rainstorm cases in Guizhou during the summer of 2020. Comparative analysis with results obtained using the MODE and neighborhood methods shows that the qualitative evaluation of prediction cases is consistent, and the performance of the SSIM* index closely aligns with the MMI index of MODE. It shows the effective application of the improved SSIM method and the new SSIM index in spatial verification for highresolution precipitation numerical prediction products, and results of the new method are even more reasonable in the assessment of prediction cases. With the whole 0-1 formatted grid fields of precipitation as the spatial objects considered, the new method can capture overall spatial structure characteristics that are not attainable through other methods, offering valuable references for assessing prediction cases based on rain area, rainband structure, and average position. Furthermore, the new method has the advantages of simple and efficient calculation, no subjective parameters setting, and well-determined results that overcome the limitations of the MODE which is the current mainstream method in spatial verification for the precipitation numerical prediction products. Therefore, this research holds good application prospects and is worthy of further study.
-
表 1 采用三种方法对2020年6月贵州区域22个暴雨个例的检验结果
日期/日 FSS MMI SSIM* 日期/日 FSS MMI SSIM* 1 0.422 0.777 0.781 14 0.460 0.814 0.611 2 0.160 0.284 0.375 18 0.516 0.258 0.685 3 0.781 0.879 0.752 21 0.499 0.789 0.632 5 0.586 0.834 0.679 22 0.688 0.879 0.760 7 0.410 0.261 0.581 23 0.422 0.862 0.763 8 0.597 0.847 0.628 24 0.582 0.802 0.787 9 0.682 0.875 0.734 25 0.487 0.255 0.507 10 0.288 0.274 0.609 27 0.481 0.856 0.683 11 0.058 0.273 0.467 28 0.901 0.859 0.930 12 0.153 0.207 0.350 29 0.105 0.273 0.576 13 0.623 0.569 0.831 30 0.494 0.826 0.781 
下载: 导出CSV
-
[1] 潘留杰, 张宏芳, 王建鹏. 数值天气预报检验方法研究进展[J]. 地球科学进展, 2014, 29(3): 327-335. [2] 陈静, 刘凑华, 陈法敬, 等. 一种基于可预报性的暴雨预报评分新方法Ⅰ: 中国暴雨可预报性综合指数[J]. 气象学报, 2019, 77(1): 15-27. [3] 陈法敬, 陈静, 韦青, 等, 一种基于可预报性的暴雨预报评分新方法Ⅱ: 暴雨检验评分模型及评估试验[J]. 气象学报, 2019, 77(1): 28-42. [4] 潘留杰, 薛春芳, 张宏芳, 等. 三种高分辨率格点降水预报检验方法的对比[J]. 气候与环境研究, 2017, 22(1): 45-58. [5] DAVIS C, BROWN B, BULLOCK R. Object-based verification of precipitation forecasts. Part Ⅰ: Methods and application to mesoscale rain areas[J]. Mon Wea Rev, 2006, 134(10): 1 772-1 784. [6] DAVIS C, BROWN B, BULLOCK R. Object-based verification of precipitation forecasts. Part Ⅱ: Application to convective rain systems [J]. Mon Wea Rev, 2006, 134(7): 1 785-1 795. [7] DAVIS C, BROWN B, BULLOCK R, et al. The method for Object-based Diagnostic Evaluation (MODE) applied to numerical forecasts form the 2005 NSSL/SPC spring program[J]. Wea Forecasting, 2009, 24(5): 1 252-1 267. [8] 刘凑华, 牛若芸. 基于对象的降水检验方法及应用[J]. 气象, 2013, 39(6): 681-691. [9] 尤凤春, 王国荣, 郭锐, 等. MODE方法在降水预报检验中的应用分析[J]. 气象, 2011, 37(12): 1 498-1 503. [10] 薛春芳, 潘留杰. 基于MODE的日本细网格模式降水预报的诊断分析[J]. 高原气象, 2016, 35(2): 406-418. [11] 赵滨, 张博. 邻域空间检验方法在降水评估中的应用[J]. 暴雨灾害, 2018, 37(1): 1-7. [12] 潘留杰, 张宏芳, 陈小婷, 等. 基于邻域法的高分辨率模式降水的预报能力分析[J]. 热带气象学报, 2015, 31(5): 632-642. [13] WANG Z, BOVIK A C, SHEIKH H R, et al. Image quality assessment: from error visibility to structural similarity[J]. IEEE Transactions on Image Processing, 2004, 13(4): 600-612. [14] WANG Z, LI Q. Information content weighting for perceptual image quality assessment[J]. IEEE Transactions on Image Processing, 2011, 20(5): 1 185-1 198. [15] 杨春玲, 徐小琳. 重视边缘区域的结构相似度图像质量评价[J]. 中国图象图形学报, 2011, 16(12): 2 133-2 139. [16] 张文娟, 张丽丽, 王艳红. 基于非局部结构张量的SSIM图像质量评价方法[J]. 计算机应用研究, 2017, 34(10): 3 162-3 164. [17] 闫乐乐, 李辉, 邱聚能, 等. 基于区域对比度和SSIM的图像质量评价方法[J]. 应用光学, 2015, 36(1): 58-63. [18] 潘旸, 谷军霞, 宇婧婧, 等. 中国区域高分辨率多源降水观测产品的融合方法试验[J]. 气象学报, 2018, 76(5): 755-766. [19] 师春香, 潘旸, 谷军霞, 等. 多源气象数据融合格点实况产品研制进展[J]. 气象学报, 2019, 77(4): 774-783. [20] 鄢俊一, 莫伟强. 2012年汛期华南中尺度模式与日本模式的降水预报评估分析[J]. 热带气象学报, 2014, 30(3): 594-600. [21] 刘段灵, 陈超. 最优TS评分方法在逐时降水订正中的初步应用及评估[J]. 热带气象学报, 2022, 38(4): 611-620. [22] 林晓霞, 冯业荣, 陈子通, 等. 华南区域高分辨率数值模式前汛期预报初步评估[J]. 热带气象学报, 2021, 37(4): 656-668. [23] 蔡利梅. MATLAB图像处理—理论、算法与实例分析[M]. 北京: 清华大学出版社, 2020. -
点击查看大图
图(5) / 表(1)
计量
- 文章访问数: 139
- HTML全文浏览量: 40
- PDF下载量: 50
- 被引次数: 0
粤公网安备 4401069904700003号