Microphysical Properties of Three Types of Snow Clouds : Implication to Satellite 1

Abstract. Ground-based radar and radiometer data observed during the 2017–18 winter were used to simultaneously estimate both cloud liquid water path and snowfall rate for three types of snowing clouds: near-surface, shallow and deep. Surveying all the observed data, it is found that near-surface cloud is the most frequently observed cloud type with an area fraction of over 60 %, while deep cloud contributes the most in snowfall volume with about 50 % of the total. The probability distributions of snowfall rates are clearly different among the three types of clouds, with vast majority hardly reaching to 0.3 mm h−1 (liquid water equivalent snowfall rate) for near-surface, 0.5 mm h−1 for shallow, and 1 mm h−1 for deep clouds. However, liquid water path in the three types of clouds all has substantial probability to reach 500 g m−2. There is no clear correlation found between snowfall rate and liquid water path for any of the cloud types. Based on all observed snow profiles, brightness temperatures at Global Precipitation Measurement Microwave Imager channels are simulated, and the ability of a Bayesian algorithm to retrieve snowfall rate is examined using half the profiles as observations and the other half as a priori database. Under idealized scenario, i.e., without considering the uncertainties caused by surface emissivity, ice particle size distribution and particle shape, the study found that the correlation as expressed by R2 between the “retrieved” and “observed” snowfall rates is about 0.33, 0.48 and 0.74, respectively, for near-surface, shallow and deep snowing clouds over land surface; these numbers basically indicate the upper limits capped by cloud natural variability, to which the retrieval skill of a Bayesian retrieval algorithm can reach. A hypothetical retrieval for the same clouds but over ocean is also studied, and a major improvement in skills is found for near-surface clouds with R2 increased from 0.33 to 0.54, while virtually no change in skills is found for deep clouds and only marginal improvement is found for shallow clouds. This study provides a general picture of the microphysical characteristics of the different types of snowing clouds and points out the associated challenges in retrieving their snowfall rate from passive microwave observations.


analyzed for snowing clouds in the following sections have a near-surface radar reflectivity 231 greater than -20 dBZ.

232
Cloud top height is used for the determination of cloud types. As shown in Fig.2,233 radar reflectivity above cloud top is often noisy as shown between 11 and 16 UTC. 234 Therefore, it is often problematic to determine cloud top height by simply using a radar 235 reflectivity threshold. However, we found that Doppler spectral width is a reliable indicator 236 to identify clouds as shown in the bottom panel in Fig.2. Using visual examination, we 237 found that Doppler spectral width commonly reduces to less than 0.1 m s -1 above cloud top.  Surveying all observed data for the entire winter, the relative frequencies of 289 occurrence (area fraction) and snowfall amount (volume fraction) for the three types of snowing clouds are shown in Fig.3. As described earlier, we used -20 dBZ at the lowest 291 bin to identify snow events. The snowfall volume is the accumulated snowfall with the rate 292 estimated by eq.(2) from radar reflectivity at the lowest bin. Over half (67.4%) of the 293 observed samples are near-surface snowfall, followed by shallow (21.2%) and then deep 294 (11.4%) snowing clouds. However, deep snowing clouds contribute the most to the total 295 snowfall volume (45.3%), followed by shallow (28.5%) and then near-surface (26.2%) 296 snowing clouds. Peninsula, and solid precipitation was observed at the radar site from 09 UTC on the 7 th 304 through 24 UTC on the 8 th . In Fig.4 shown are cross section of radar reflectivity and time 305 variation of liquid water path and snow water path (SWP, vertically integrated snow water 306 content). Surface PARSIVEL and 2DVD observations indicated that snow particle types 307 are mostly snowflakes from 09 UTC on the 7 th to 06 UTC on the 8 th , while rimed ice 308 particles and graupels are also observed then after. The radar and radiometer observations 309 indicate that the deep clouds have cloud top higher than 8 km and peak snow water path 310 value about 400 g m -2 . However, liquid water in the deep clouds is low, with liquid water 311 path constantly below 150 g m -2 . Once the deep clouds pass the station, the clouds became 312 much shallower, mostly being classified as near-surface snowing clouds. However, their 313 liquid water path increased substantially with peak values close to 600 g m -2 , which is 314 consistent with the observed rimed ice particles and graupels during this period.  implying that most of the precipitation growth occurs above 3 km. For shallow clouds, the 385 "left-tilting" structure starts from near surface and the frequency has broader distribution 386 at each level. In contrast, the near-surface snowing clouds do not show such "left-tilting" 387 structure, but rather have a broad distribution below their cloud top height, indicating that 388 the precipitation maximum does not necessarily situate near the surface in these profiles. 389 We interpret that the broad distribution of frequencies at each level is likely due to the 390 convective nature of these clouds, so that the precipitation profile is largely determined by    Figure 11 shows how brightness temperature varies as liquid water path and surface 426 snowfall rate increase. Note that in these calculations, we used the observed snowfall rate 427 profiles derived for each cloud type and averaged for various snowfall rate bins. A 1-km 428 deep liquid cloud layer is placed at 0.5-1.5 km, 2.5-3.5 km and 4.5-5.5 km, respectively, 429 for near-surface, shallow, and deep clouds. The liquid water path is increased from 0 to 430 500 g m -2 .

431
For near-surface snowing clouds, the decrease of brightness temperature due to ice 432 scattering is very limited for either 89 or 166 GHz, only a few Kelvin occurring when liquid 433 water path is very low. Therefore, most likely this type of clouds displays a warming 434 signature in the passive microwave observations due to the existence of liquid water clouds. the same clouds were moved to areas over ocean where liquid water information is 498 distinguishable at some microwave channels (e.g., 89 GHz)? To answer this question, we 499 perform the same retrieval exercise as mentioned above but assuming the clouds are over  Fig.13. For deep snowing clouds, the R 2 statistic indicates virtually no difference 503 in retrieval skills between over land and over ocean cases, although a visual inspection of 504 the scatterplot shows that a better correspondence between "measured" and retrieved 505 values at snowfall rates lower than 0.2 mm h -1 . The improvement in retrieval skills for over 506 ocean shallow clouds is marginal with R 2 of 0.54 versus 0.48 over land. The most 507 significant improvement in retrieval skills occurs for over ocean near-surface snowing 508 clouds, in which R 2 increases from 0.33 over land to 0.54 over ocean. Note that land surface 509 emissivity and ocean surface wind are fixed in the retrieval exercises. Therefore, the 510 improvement is not due to a better knowledge of surface conditions, but rather due to the 511 richer information content on cloud microphysics contained in "measured" brightness 512 temperatures over ocean. One such piece of information must have come from the 513 brightness temperature difference between two polarizations over ocean, which remines 514 mostly zero over land surfaces. The results shown in Fig.13 indicate that the extra 515 polarization information helps the most for retrieving snowfall in near-surface clouds.

516
To understand the information conveyed in polarization difference of brightness 517 temperatures, we performed a similar simulation to that described in Section 4.1, but 518 replaced land surface to ocean surface with a wind speed of 5 m s -1 . The changes of 519 depolarization as liquid water path and snowfall rate increase are shown in Fig.14  Therefore, it is plausible that the increased retrieval skill over ocean for near-surface and 526 shallow clouds is due to the added information on liquid water contained in the polarization differences. Comparing Figs.12 and 13, it seems that the added information is particularly 528 helpful in improving retrievals at low snowfall rates.  Surveying all the observed data, it is found that near-surface snowing cloud is the most 551 frequently observed cloud type with a frequency of occurrence over 60%, while deep 552 snowing cloud contributes the most in snowfall volume with about 50% of the total 553 snowfall amount.

554
The probability distributions of surface snowfall rates are clearly different among 555 the three types of snowing clouds, with vast majority of it hardly reaching to 0.3 mm h -1