| Along with the applications of various advanced meteoroligcal satelliate platforms and advancements of remote sensing techniques, the macro and micro characteristics of clouds and precipitation, as well as the roles played by them in the climate system are revealed step by step. Analyses of clouds and precipitation based on the joint observations of multi-instrument and multi-channel observations has been an active research field in atmospheric remote sensing and climate change. In this study, based on the TRMM PR and VIRS measurements, which provide high-quality precipitation measurements and refined radiative information from cloud top, together with reliable cloud detection and cloud parameters retrieval techniques, the climatological characteristics of precipitating clouds (PCs) and non-precipitating clouds (NPCs) was investigated within a long time and a global scale. Three daytime precipitating clouds identification schemes from cloud parameters information have been proposed, and the relationships between cloud parameters and precipitation have been studied. Particularly, the differences of cloud parameters between PCs and NPCs were emphasized. The preliminary results on three aspects are presented as following.(1) Daytime PCs identification scheme from cloud parameters informationBased on synchronous precipitation and radiative measurements, the precipitation information derived from PR and the cloud parameters retrieved from VIRS were compared and the relationships between cloud parameters and precipitation probability were analyzed. Three frameworks for determining the occurrence of precipitation in daytime, called Identification of Precipitating Clouds from Optical Thickness and Effective Radius (IPC rRe), Identification of Precipitating Clouds from Thermal Infrared Brightness Temperature and Cloud Parameters (IPCTC), Precipitation Identification Scheme from Cloud Parameters information (PISCP), respectively, were developed.It was proved that PISCP performed better than IPCTC and IPC τRe both in land and oceanic areas. The performance of IPCTC was followed by PISCP, and IPC τRe a performed worse than PISCP and IPCTC. Over land area, the equitable threat score (ETS) values of PISCP, IPCTC and IPC τRe were0.37,0.36and0.34, respectively. The performances of the three schemes in oceanic areas were all better than land areas, and the ETS values were0.44,0.42and0.40.Compared with existing similar precipitation identification schemes, the accuracies of the three proposed schemes were better, especially of PISCP, which was close to the accuracy of passive microwave instruments. The results also showed that the three proposed schemes cloud widely applied in other regions of the world.(2) Relationships between cloud parameters and precipitation in rainstorm areasFocusing on a set of summer rainstorm samples in eastern China, the synchronous precipitation profiles and cloud parameters datasets were developed, and the relationships between cloud parameters and precipitation were analyzed, especially the relationships between cloud parameters and precipitation vertical structures.Results indicated that the ratio of ice-cloud in PCs (-87%) was much higher than in NPCs (-40%). Both optical thickness (r) and cloud water path (CWP) of PCs were obviously larger than that of NPCs, and the cloud top temperature (Tc) were lower but the differences in effective radius (Re) were very small. As r and CWP increased (Tc decreased), the probability of precipitation increased. Among these, the CWP and Tc have obvious effects in discriminating precipitation, and the effects of Re was weak.Whether stratiform precipitation or convective precipitation, the r, CWP and Re increased with increasing of near surface rainrate, meanwhile the Tc decreased. For a given near surface rainrate, the rand CWP of convective precipitation were smaller than those of stratiform precipitation, but the Re of them were exactly the same.As cloud parameters changing, the response characteristics of precipitation profiles to were similar for stratiform precipitation and convective precipitation. Specifically, with r, Re and CWP increasing and Tc decreasing, the rain top heights decreased and the rainrates lower than5km consistently decreased. Only Re less than15μm, the precipitation profiles have notable changes. For convective precipitation, raindrops of4km to2km levels evaporated significantly when CWP were low.(3) Differences of cloud parameters between PCs and NPCsThe spatial distribution characteristics of PCs and NPCs were analyzed on climate scale. Combined with NCEP reanalysis datasets, the key meteorological elements of the decision to spatial distribution of PCs were studied. We Focused on comparison the differences of cloud parameters mean value and spatial distribution between PCs and NPCs.Results showed that the cloud amounts in the high precipitation frequency area were generally higher than80%over global scale. However, the precipitation frequencies in the high high cloud amounts areas were not necessarily high, some of these areas even hardly occurred precipitation. The proportion of stratiform precipitation was highest (～79%), convective precipitation followed by stratiform precipitation (～21%), and the proportion of other-type precipitation was lowest (～0.5%). The global average daily rainfall amount was about2.6mm/d. The rainfall contributions of stratiform precipitation and convective precipitation were near, which were53%and47%, respectively. While the contribution of other-type precipitation was very small, which was less than0.1%.Compared with NCEP reanalysis datasets, it was found that the high/low value areas of relative humidity at middle level (600hPa) were highly corresponded to the high/low value areas of total precipitation frequency. The latitudinal correlation coefficient between them was0.90, and longitudinal correlation coefficient was0.60. The divergence at high level (200hPa) was more consistent corresponding with total precipitation frequency, and the latitudinal (longitudinal) correlation coefficient between them was up to0.98(0.88). These showed that the planetary-scale water vapor transports and atmospheric circulations were the dominant factors of global total precipitation frequency distributions, in which large-scale atmospheric rising/sinking was the decisive factor.Results indicated that most of the low cloud precipitation was shallow, isolated precipitation, and the proportion was more than60%. The precipitation intensity and spatial scales of these shallow, isolated precipitation were small. Results also showed that the difference distribution of vertical velocity between middle (600hPa) and high (200hPa) levels corresponded with the precipitation frequency distribution of middle cloud precipitation, and the latitudinal (longitudinal) correlation coefficient between them was0.61(0.51). The vertical velocity distribution of high (200hPa) levels corresponded well with the precipitation frequency distribution of high cloud precipitation, and the latitudinal (longitudinal) correlation coefficient between them was up to0.94(0.83). These results indicated that atmospheric rising/sinking at middle (high) level was the dominant factor of middle (high) cloud precipitation frequency distribution.On a global scale, the cloud amounts of NPCs were much higher than PCs, and the ratio between them was more than5times. The high cloud amount areas of NPCs and PCs were not entirely consistent, but the low cloud amount areas of them were the same. The statistical results also showed that with the cloud top height increased, the cloud amount ratios between NPCs and PCs decreased.Results showed that the differences of τ between PCs and NPCs were very significant, which were larger than60, and the spatial distribution of τ differences between PCs and NPCs was similar to the spatial distributions of precipitation frequency and averaged daily rainfall amount. The Re of PCs were significantly larger than NPCs in oceanic areas, but no obvious difference in land areas. The segment statistics showed that the Re of PCs were larger than NPCs in different τ segments. The no obvious Re difference in land areas was due to the different τspectrums of PCs and NPCs.Analysis also showed that the differences in CWP between PCs and NPCs were obviously larger than the differences of both τ and Re. The Tc of PCs were significantly lower than NPCs, and the spatial distributions of cloud top pressure and cloud top height were similar to Tc. The statistics also showed that the proportion of ice cloud in PCs (～76%) was much higher than that in NPCs (～23%). |
PDF Full Text Request