Characterizing the near-global cloud vertical structures over land using high-resolution radiosonde measurements
Abstract. Cloud remains one of the largest uncertainties in weather and climate research due to the lack of fine-resolution observations of cloud vertical structure (CVS) on large scale. In this research, near-global CVS is characterized by high-vertical-resolution twice daily radiosonde observations from 374 stations over land. It is found that the cloud base heights (CBHs) from the radiosondes have a higher correlation coefficient (R = 0.91) with the millimeter wavelength cloud radar than that with the ERA5 reanalysis (R = 0.49). Overall, cloudy skies occur 65.3 % (69.5 %) of the time, of which 55.4 % (53.8 %) are one-layer clouds at 0000 (1200) UTC. Most multi-layer clouds are two-layer clouds, accounting for 62.2 % (61.1 %) among multi-layer clouds for 0000 (1200) UTC. Geographically, one-layer clouds tend to occur over arid regions, whereas two-layer clouds do not show any clear spatial preference. The cloud bases and tops over arid regions are higher compared with humid regions albeit with smaller cloud thickness (CT). Clouds tend to have lower bases and thinner layer thicknesses as the number of cloud layer increases. The global mean CT, CBH, and cloud top height (CTH) are 4.89 ± 1.36 (5.37 ± 1.58), 3.15 ± 1.15 (3.07 ± 1.06), and 8.04 ± 1.60 (8.44 ± 1.52) km above ground level (AGL) at 0000 (1200) UTC, respectively. The occurrence frequency of clouds is bimodal with lower peaks between 0.5 and 3 km AGL and upper peaks between 6 and 10 km AGL. The CBH, CTH and CT undergo almost the same seasonality that their magnitudes are greater in the boreal summer than in the winter. As expected, the occurrence frequencies of clouds exhibit pronounced diurnal cycles in different seasons. In boreal summer, clouds tend to form as sun rises and the occurrence frequencies increase from morning to later afternoon, with the peak in the early afternoon at altitudes 6–12 km; while in boreal winter, clouds have peak occurrence frequencies in the morning. The relations between surface meteorological variables and moisture with CBH are investigated as well, showing that CBH are generally more significantly correlated with 2 m RH (RH2m) and 2 m T (T2m) than with surface pressure and 10 m wind speed. Larger T2m and smaller RH2m always correspond to higher CBH. In most cases CBHs are negatively correlated to soil water content. The near-global CVS obtained from high-vertical-resolution radiosonde in this study can provide key data support for improving the accuracy of cloud radiative forcing simulation in climate models.
Hui Xu et al.
Status: open (until 10 Jun 2023)
- RC1: 'Comment on egusphere-2023-472', Anonymous Referee #1, 20 May 2023 reply
Hui Xu et al.
Hui Xu et al.
Viewed (geographical distribution)
This paper examines the near-global cloud vertical structures using two years of radiosonde data. I do not find any major flaws with their methodology and conclusions, and the statistical results could be a nice contribution to modeling global cloud radiative effects. However, clarifications are needed to make this paper a compelling story. I suggest returning to the authors for minor revision.
The Introduction section listed several previous works using lowering resolution radiosonde data to retrieval cloud boundaries but did not include a summary of what were found from those works, what are the main statistical and conclusions from those works. Most importantly, the authors should articular what are novel in the current study, in addition to higher resolution data.
Line 17-19: cloud base height correlate with millimeter wavelength radar?
Line 52: do you mean the Chang and Li retrievals have large discrepancies? Discrepancies relative to what?
Line 55-56: the last sentence needs to be revised. Polar orbiting satellites can have short revisit periods such as AQUA/TERRA. Do you mean ‘narrower nadir views’ ?
Line 58: cloud radars
Line 75: do you mean the vertical resolution, horizontal resolution, or temporal resolution?
Line 75-79: it will be more intuitive to understand the difference of ‘resolution’ (whatever it refers to) from previous and current radiosondes if you can provide several numbers here.
Line 107: change ‘considered’ to ‘included’
Line 115: an accuracy of
Line 124-125: references for the ERA5 reanalysis are needed here
Line 168: enters a moist layer
Line 190: can you explain why a max-RH is needed to detect a cloud layer? What is inter-RH in Table 1 and Figure 2? Is it the RH between consecutive cloud layers?
Line 184-191: do you do any averaging or smoothing for the RH and T profiles, considering they are in high vertical resolution?
Figure 3: I suggest change sounding times to 00UTC and 12UTC to be consistent with your intro text
line 223: maybe change the word ‘correctly’ to ‘reasonably’
line 313-314: these result in the occurrence
line 368: oceanic climate