JpGU-AGU Joint Meeting 2017

講演情報

[EE]Eveningポスター発表

セッション記号 A (大気水圏科学) » A-AS 大気科学・気象学・大気環境

[A-AS08] [EE] 雲降水過程の統合的理解に向けて

2017年5月23日(火) 17:15 〜 18:30 ポスター会場 (国際展示場 7ホール)

[AAS08-P03] Estimating vertical profile of water-cloud droplet effective radius from SWIR measurements of Himawari-8 via cloud profile statistics

*永尾 隆1中島 孝2 (1.宇宙航空研究開発機構地球観測研究センター、2.東海大学情報技術センター)

Remote sensing of clouds by geostationary meteorological satellites with multispectral visible-to-infrared imaging capabilities by improved spectral, spatial and temporal resolution (e.g. Himawari-8/AHI, GOES-R/ABI) have potential to advance scientific understanding of cloud and precipitation process, by quantifying spatio-temporal distributions and evolutions of cloud radiative and microphysical properties such as cloud optical thickness (COT), cloud droplet effective radius (CDER), and cloud top temperature (CTT) and height (CTH). However, limitation of passive remote sensing to provide vertically-resolved cloud information including in-cloud CDER vertical profile (CDER-VP), drizzling existence, cloud geometric thickness and base height is one of the barriers to advancing our understanding of cloud and precipitation process.
This study developed an algorithm to retrieve CDER-VP of water cloud from shortwave infrared (SWIR) measurements of Himawari-8/AHI via cloud statistical profiles derived from CloudSat/CPR observation towards continuous monitoring of temporal evolution of clouds by Himawari-8. Several similar algorithms in previous studies utilize a spectral radiance matching on the assumption of simultaneous observation of radar and visible-to-infrared imagery such as CloudSat/CPR and Aqua/MODIS. However, note that, since our aim is to apply the algorithm to Himawari-8/AHI measurements, the algorithm does not assume simultaneous observations with CloudSat/CPR.
First, in advance, a database (DB) of CDER-VP is prepared by the following procedure: Top-of-atmosphere (TOA) radiances at three spectral bands (0.65, 2.3, 11-μm) of AHI are simulated from CDER-VP contained in CloudSat Radar-Visible Optical Depth Cloud Water Content Product (2B-CWC-RVOD) and cloud optical depth vertical profile in contained in CloudSat 2B-TAU product. COT, CDER and CTT are retrieved from the simulated radiances using an algorithm in assuming plane-parallel cloud structure (Nakajima and Nakajima, 1995). A set of the COT, CDER and CTT retrievals and inputted CDER-VP is added to the DB. Finally the algorithm retrieves CDER-VP from actual AHI measurement by the following procedure: COT, CDER and CTT are retrieved from the AHI radiances at 0.64, 2.3, 10.4-μm bands. Using the COT, CDER and CTT retrievals as key of the DB, multiple CDER-VPs are extracted from the DB. Using principal component (PC) analysis, up to three PC vectors of the CDER-VPs are extracted. Again, CDER-VP, COT, and CTT are retrieved from the AHI 0.64, 1.6, 2.3, 3.9 and 10.4-μm using the PC vectors of CDER-VPs with iterative radiative transfer calculation. Note that the sum of the contribution ratios of the three principal components vectors exceeds 95%.
This study evaluated the algorithm based on the simulation using the CloudSat 2B-CWC-RVOD and 2B-TAU products. The column mean CDERs calculated from the retrieved CDER-VPs are almost unbiased while the CDERs retrieved with the plane-parallel assumption have significant underestimation called so-called plane-parallel bias. The CDER-VP retrieval errors are almost smaller than 3-μm. The retrieval errors of the cloud base CDERs are almost larger than the others. The tendency can be explained by less sensitivity of SWIRs to CDER at cloud base.
Additionally, as a case study, this study will attempt to apply the algorithm to the AHI’s high-frequency observations, and to interpret the time series of the CDER-VP retrievals in terms of temporal evolution of water clouds.