GPM/DPR雷达与CINRAD雷达降水探测对比

Comparison on the Precipitation Measurement Between GPM/DPR and CINRAD Radars

  • 摘要: 通过对比星载DPR雷达与地基CINRAD雷达的降雨测量值,评估星地雷达联合应用的潜力。为了提高对比的准确性,在尽可能高的时空分辨率下,以几何匹配与格点匹配相结合的方式,提取星地雷达降水样本数据。2015年6月30日降水过程的对比分析结果表明:泰州、常州CINRAD雷达反射率因子在两站中分剖面的平均值偏差0.94 dB,地基雷达之间有很好的一致性;在DPR雷达与常州、泰州CINRAD雷达同时覆盖的降雨区域,星地之间雷达反射率因子的平均值偏差分别为-1.2 dB和-1.6 dB,显示星地雷达也有较好的一致性;现有DPR雷达陆上衰减订正算法在缩小星地雷达偏差方面起到一定作用,平均订正量0.4 dB,只要回波覆盖充分,匹配样本的高度以及其到地基雷达的距离对对比结果没有明显影响,而衰减订正和匹配样本区回波覆盖率是影响星地雷达对比结果的重要因素。

     

    Abstract: It is necessary to find out the difference between space-borne and ground-based radar data for evaluating the possibility of combined use of them. 2 neighboring ground-based radars at Taizhou and Changzhou are first checked for data consistence in full resolution and then compared with the DPR radar respectively. Results from the precipitation case on 30 June 2015 show that 2 radars have 0.94 dB bias of mean reflectivity factor on the profile where distances to both radars are equal.To get high temporal and spatial resolution comparisons between DPR and CINRAD, the geometry-matching algorithm is used in vertical where 1-14 DPR range gates could be included in one sample pairs according to the distance to CINRAD radar site. The further away it's from the radar site, the more DPR gates can be included. The grid-matching algorithm is used in horizontal where total 5×5 grids in 1 km resolution are matched with one single DPR range gate. The DPR and CINRAD volume-averaged values are calculated for all such intersecting DPR range gates and 5×5 CINRAD grids. Statistic results on sample pairs show that the mean reflectivity factor biases of DPR radar are -1.2 dB and -1.6 dB for CINRAD Taizhou and Changzhou radars, respectively, and the mean rain rate converted from Z-R relationship are 0.10 mm·h-1 and 0.13 mm·h-1 lower than CINRAD radars', over the same area where the three radars scanned successively within 6 min. When the distance to CINRAD gets longer, the bias between DPR and CINRAD is larger near the top of echo. And the bias in the bright band area is 122 dB larger than the mean bias as well. But the bias has no obvious relevance with distance and height in the other area if beam filling is enough.Attenuation correction and echo coverage over sample cell are among important factors which affect comparison results. Though there is no suitable surface reference, the attenuation correction algorithm for DPR radar over land works and decreases 0.4 dB in mean bias between DPR and CINRAD Taizhou. The maximum correction is only 1.36 dB due to moderate intensity of radar reflectivity factor.The equivalent radar reflectivity factor must be modified for the application of comparing DPR radar to other wavelength radar when the equivalent radar reflectivity factor is greater than 37 dBZ. For the application of combined multi-wavelength (such as DPR and CINRAD), data quality control and clutter identification and elimination are all among direct acting factors.

     

/

返回文章
返回