the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
Impacts of combined microphysical and land-surface uncertainties on convective clouds and precipitation
- 1Institute of Meteorology and Climate Research (IMK-TRO), Department Troposphere Research, Karlsruhe Institute of Technology (KIT), Karlsruhe, Germany
- 2Meteorologisches Institut, Ludwig-Maximilians-Universität, Munich, Germany
- 1Institute of Meteorology and Climate Research (IMK-TRO), Department Troposphere Research, Karlsruhe Institute of Technology (KIT), Karlsruhe, Germany
- 2Meteorologisches Institut, Ludwig-Maximilians-Universität, Munich, Germany
Abstract. To reduce the underdispersion of precipitation in convective-scale ensemble prediction systems, we investigate the relevance of microphysical and land-surface uncertainties for convective-scale predictability. We use three different initial soil moisture fields and study the response of convective precipitation to varying cloud condensation nuclei (CCN) concentrations and different shape parameters of the cloud droplet size distribution (CDSD) by applying a novel combined-perturbation strategy. Using the new icosahedral nonhydrostatic ICON model, we construct a 60 member ensemble for cases with summertime convection under weak and strong synoptic-scale forcing over central Europe. We find a systematic positive soil moisture–precipitation feedback for all cases, regardless of the type of synoptic forcing and a stronger response of precipitation to different CCN concentrations and shape parameters for weak forcing than for strong forcing. While the days with weak forcing show a systematic decrease in precipitation with increasing aerosol loading, days with strong forcing also show nonsystematic responses for some values of the shape parameters. The large magnitude of precipitation deviations compared to a reference simulation ranging between −23 % to +18 % demonstrates that the uncertainties investigated here and, in particular, their collective effect are highly relevant for quantitative precipitation forecasting of summertime convection in central Europe. A rain water budget analysis is used to identify the dominating source and sink terms and their response to the uncertainties applied in this study. Results also show a dominating cold-rain process for all cases and a strong, but mostly non systematic impact on the release of latent heat which is considered to be the prime mechanism for the upscale growth of small errors affecting the predictability of convective systems. The combined ensemble spread when accounting for all three uncertainties is generally larger for weak forcing cases and lies in the same range than the ones from an operational convective-scale ensemble prediction system determined in previous studies. This indicates that the combination of different perturbations used in our study may be suitable for ensemble forecasting and that this method should be evaluated against other sources of uncertainty.
Christian Barthlott et al.
Status: open (until 13 Jul 2022)
-
RC1: 'Comment on acp-2022-322', Anonymous Referee #2, 17 Jun 2022
reply
As the authors stated, model uncertainties exist in many physical schemes and dynamics, including initial/boundary conditions (IC/BCs). There can be numerous combinations of such uncertainty sources. Can the authors explain why it is essential to consider the combination of soil moisture and cloud microphysics compared to other possible combinations?
A similar question: there are many uncertainties in cloud microphysical parameterizations. How are these factors considered in this study? Can the authors justify why they focused only on uncertainties in N_CCN and CDSD parameters? Also, the authors mentioned many uncertainties related to aerosol-cloud interactions (lines 45-76). Can these uncertainties be represented by perturbing the N_CCN?The discussion on the normalized standard deviation is too brief and does not provide much scientific insight.
I would like to see a more quantitative comparison of spreads from the three sensitivity factors (i.e., soil moisture, CCN, and shape factor.
-
RC2: 'Comment on acp-2022-322', Anonymous Referee #2, 17 Jun 2022
reply
This study investigated the model uncertainties associated with three factors: soil moisture, CCN concentration, and the shape factor of cloud drop size distribution. Quite a few similar studies have been conducted recently, but none applied such a three-parameter combination. The results showed significant spreads caused by the two microphysical parameters, and the soil moisture factor also enhances the spread. But the significance of these factors compared to many others in the model is unclear. The manuscript can be enriched if the suggestions in the major comments below can be considered.
Note: I have posted a preliminary version of the review, and sorry for double-posting some of the comments.
Major comments
Introduction and methodology:
1. As the authors stated, model uncertainties exist in many physical schemes and dynamics, including initial/boundary conditions (IC/BCs). There can be numerous combinations of such uncertainty sources. Can the authors explain why it is essential to consider the combination of soil moisture and cloud microphysics compared to other possible combinations?
2. Similarly, there are many uncertainties in cloud microphysical parameterizations. How are these factors considered in this study? Can the authors justify why they focused only on uncertainties in NCN and CDSD parameters? Also, the authors mentioned many uncertainties related to aerosol-cloud interactions (lines 45-76). Can these uncertainties be represented by perturbing the NCN?
3. Line 122-123: There is a difference between NCN and NCCN (CN stands for condensation nuclei and CCN for cloud condensation nuclei). For polluted continental conditions, the value of 3200 cm-3 seems to be too low for NCN (should be tens of thousands or more) but fine for NCCN. The values used for other conditions should also be justified or, at least, provide a reference.
4. The shape parameter ν is also important for other hydrometeors. In fact, the variation in ν may be even more prominent for precipitation particles according to some triple-moment schemes. What is the reason for perturbing only ν of cloud drops?Results:
1. The model “spread” is one of the key foci of this study. Yet, the discussion on the normalized standard deviation (indicating the spread) is too brief and does not provide much scientific insight.
2. I would like to see a more quantitative comparison of spreads from the three sensitivity factors (i.e., soil moisture, CCN, and shape factor. This allows the reader to judge which factors are more important for the consideration of ensemble members.
3. Line 249-250: “The higher the CCN concentration, the lower are the rain intensities.”
This seems to be a warm-rain characteristic. But, apparently, the studied systems are mostly cold-rain dominant (lines 443-444). For mixed-phase convective systems, higher aerosol concentrations often lead to stronger rain intensity (cf. Tao et al. 2012, etc.). It will be nice to compare the results here with other relevant studies.
4. Figure 7.
The tendency of TQG change with ν is different for maritime CN compared to other CN types for cases 2018 and 2020. Some inconsistencies also exist in the 2016 case. Is there any explanation?Conclusion:
1. It is dangerous to make a conclusion based on only four cases. Large differences can be observed between the two weak cases or between the strong cases, which may suggest that other cases may behave distinctively differently and even produce results that disagree with the conclusions stated here. Furthermore, the uncertainties in the studied parameters may vary if you choose different initial/boundary conditions, physics schemes, or grid resolutions. The authors should at least try to tone down a bit on the certainty of their findings.
2. Perhaps the authors can make a quantitative comparison of the spread caused by each factor by preparing a table summarizing the relative spreads (standard deviation).
3. There are quite a few similar studies with multiple-factors analyses. Because of the numerous possible combinations of uncertainty factors, it will be nice to see some comparisons on the spread/uncertainty with previous studies.Minor comments
1. Line 5: 60 member ensemble ⇒ 60-member ensemble (same in other places of the text)
2. Line 12-13: rain water ⇒ rainwater
3. Line 14: strong, but ⇒ strong but
4. Line 14: non systematic ⇒ non-systematic
5. Line 15: which ⇒ , which
6. Equation (1): Since the microphysics scheme used is double moments with A and λ as varying coefficients, ν and μ must be specified. The value for μ was never mentioned. If μ was set to 1, then just omit it in the equation.
7. Figures 5, 7-9: These figures are quite complicated. More details (e.g., what is NU) are needed in the caption to assist the readers in understanding the arrangements.
8. Line 215: applies for ⇒ applies to -
RC3: 'Comment on acp-2022-322', Anonymous Referee #1, 26 Jun 2022
reply
This paper examined the precipitation and updraft changes over Germany and parts of neighboring countries to four different parameters: CCN concentration, shape parameter, soil moisture, and synoptic forcing.
The first result section is a series of descriptions of what's shown in the simulations rather than a coherent synthesis/comparison. The reviewer strongly suggests that the authors develop a better way of organizing and displaying the results, particularly for the precipitation responses to three perturbed parameters. The authors need to separate materials into different subsections focusing on one perturbed parameter at a time and demonstrating the synergistic effects explicitly. Furthermore, the results are discussed when one of four parameters is constrained. For example, Figure 6 and associated texts illustrate the impact of soil moisture under different synoptic setups. As such, examination of the synergistic/interactive effects of more than two perturbed parameters is absent, which is critical in terms of adding novel findings to the existing literature on modulation of convective responses by interactions between aerosol and soil moisture (or other environmental parameters). Addressing the synergistic is imperative since the title is Impacts of "combined" microphysical and land-surface uncertainties. Also, since the simulation results are quite different depending on the synoptic setting, the reviewer recommends authors to revise the title to include the synoptic aspect.
Finally, the third result section only examines the updraft at 5 km, which is insufficient information to examine/determine convective invigoration. The authors need to consider expanding the updraft analysis to different levels or exclude this section from the paper. The following texts are point-by-point comments.
Table 1: What is the vertical resolution? Is it varying or constant throughout the column? If varying, what is the vertical resolution near the surface, which could impact the boundary layer processes.
Lines 121–123: Are these CCN concentrations based on the observation? The reasoning for selecting these numbers should be demonstrated clearly with appropriate references.
Line 156: maritime, intermediate, continental, and polluted are not intuitive and misleading unless these CCN values are based on observations over the maritime and continental part/airmass over the simulation domain. I suggest naming them CCN100, CCN500, CCN1700, and CCN3200, for example. If this is too long, please come up with something more informative. It's tough to remember whether continental is lower or higher than polluted. All the values are different degrees of pollution/aerosol loading.
Lins 168: Could you add one or more sentences explaining the definition of convective adjustment time scale?
Table 2: The authors need to include the justification for choosing 0.17 and 1.09 as weak strong forcing cases. Unlike weak forcing cases, where the convective adjustment time scales have only ~ 11% difference, the two numbers chosen for the strong instances have more than 146% difference. These could have impacted the dramatic deviation from the reference shown in Figure 5b, first panel. A similar concern goes for Figure 6, rows 3 and 4, where the diurnal behavior looks pretty different between these two forcings compared to similar peaks and shapes shown in rows 1 and 2.
Line 181: While the agreement between simulated precipitation and radar observations is not shown here, could you elaborate more regarding what radar observations were used and what parameter (e.g., accumulated rainfall, hourly precipitation rate) was chosen to make this comparison?
Line 188: Why did you choose "domain-integrated precipitation totals" to represent the precipitation response? The black boxed area in Figure 1 includes land with substantially different orography (the south edge of the domain and the north edge of the domain), water, and coastal regions, which all could show very different precipitation characteristics. Especially under weak synoptic scenarios, the coastal rain process could impact the domain-integrated precipitation.
Lines 191–195: Was this soil moisture response linear? Several pieces of literature have shown the nonlinear characteristics of the soil moisture impact on convection. This is also related to the limitation of the current study's design. I understand the computation limitation with many simulations considering the non-linearity of soil moisture and other parameters. However, the authors should demonstrate discussion on nonlinear responses found either in this study or in previous studies more clearly in this manuscript. Please check Drager et al. (2022) "A Non-Monotonic Precipitation Response to Changes in Soil Moisture in
the Presence of Vegetation" for this comment.Lines 218–219: These referenced studies all used the COSMO model. And the paper you cited earlier, Marniescu et al. (2021), showed the different convective responses to enhanced aerosol loading resulting from different models. So the authors need to look into other papers that used the various numerical models and examined aerosol-induced convection changes under different synoptic setups.
Lines 231–234: Is there any way to show this using figures? For example, changes in instability between different simulations? While the explanation authors put here makes sense, it's merely speculation without supporting simulated results. The same goes for Lines 236–244. Where is the supporting evidence for cloud size changes?
Figure 6: Color shadings, instead of colored lines, are hard for readers to compare responses among different aerosol loadings. Please consider other ways of representing four aerosol loadings for clarity.
Line 256: Since Figure 6 only shows the shape parameter-averaged response, the (WETp0 minus WETp8) is not demonstrated via any figure or table. Could you also consider including different responses as a function of the shape parameter?
Line 284: Please also consider including other references not involving the first author.
Lines 363–365: Please explain why the authors examined this ratio. How does this ratio tell about latent heat release/updrafts? This ratio seems only relevant to the relative dominance of cold rain.
Line 377: the sensitivity to different shape parameters -> the sensitivity of w 5km(?) to different shape parameters; since there are sensitivities in the first and second rows as a function of the shape parameter.
Lines 383–384: Please include a table or a figure showing the mean updrafts to support these statements. At least, the authors should include the numbers of mean updraft value ranges or distribution.
Lines 390–391: Do you mean there is no evidence of convective invigoration (or suppression) throughout all vertical levels? No warm-phase invigoration either? Several recent studies (e.g., Igel and van den Heever, 2021 and references therein) have shown the robustness of the warm-phase invigoration, whereas the cold-phase invigoration is not robust but depends on the environment.
Line 400: Please check Grant and van den Heever (2014) for how they computed synergistic effects when multiple parameters were perturbed.
Christian Barthlott et al.
Christian Barthlott et al.
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 225 | 30 | 9 | 264 | 1 | 0 |
- HTML: 225
- PDF: 30
- XML: 9
- Total: 264
- BibTeX: 1
- EndNote: 0
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1