the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
Understanding the dependence of mean precipitation on convective treatment in tropical aquachannel experiments
Peter Knippertz
Yvonne Ruckstuhl
Robert Redl
Tijana Janjic
Corinna Hoose
Abstract. The intertropical convergence zone (ITCZ) is a key circulation and precipitation feature in the tropics. There has been a large spread in the representation of the ITCZ in global weather and climate models for a long time, the reasons for which remain unclear. This manuscript presents a novel approach with which we disentangle different physical processes responsible for the changeable behavior of the ITCZ in numerical models. The diagnostic tool is based on a conceptual framework developed by Emanuel (2019) and allows for physically consistent estimates of convective mass flux and precipitation efficiency for simulations with explicit and parameterized convection. We apply our diagnostics to a set of tropical aquachannel experiments using the ICOsahedral Nonhydrostatic (ICON) model with horizontal grid resolution of 13 km and with various representations of deep and shallow convection. The channel length corresponds to the Earth's circumference and has rigid walls at 30° N/S. Zonally symmetric sea surface temperatures are prescribed.
All four runs share overall similar rainfall patterns and dynamical structures. They simulate an ITCZ at the equator coinciding with the ascending branch of the Hadley circulation, descending branches at 15° N/S with subtropical jets and easterly trade wind belts straddling the ITCZ. Differences are largest between runs with and without parameterized deep convection. With explicit deep convection, rainfall in the ITCZ increases by 35 % and the Hadley circulation as well as surface winds become stronger. Our diagnostic framework reveals that boundary-layer quasi-equilibrium is a key to physically understanding those differences. The stronger surface horizontal winds with explicit deep convection essentially enhance surface enthalpy fluxes and thus perturb quasi-equilibrium in the boundary layer. This is balanced by increasing convective downdraft mass flux that carries low moist static energy from the lower troposphere into the boundary layer. The downdraft strength is proportional to convective updraft mass flux, which is closely linked to rainfall, since – somewhat surprisingly – the convective treatment does not appear to influence precipitation efficiency significantly. Changes in radiative cooling are largely compensated by changes in dry stability, leading to little impact on rainfall. The results highlight the utility of our diagnostics to pinpoint processes important for rainfall differences between models, suggesting applicability for global climate model intercomparison projects.
- Preprint
(1728 KB) - Metadata XML
-
Supplement
(88513 KB) - BibTeX
- EndNote
Hyunju Jung et al.
Status: final response (author comments only)
-
RC1: 'Comment on wcd-2023-7', Anonymous Referee #1, 01 Mar 2023
This study investigates the impact of convective treatment (particularly, parameterized vs explicit deep convection) on the simulation of mean tropical precipitation (particularly, the ITCZ), using a 13km ICON model in a semi-aquaplanet simulation with walls at 30N/S. They find, with explicit deep convection, rainfall in the ITCZ increases by 35% and the Hadley circulation as well as surface winds become stronger. Based on a diagnostic framework based on Emanuel (2019), they attribute the difference to the stronger surface horizontal winds with explicit deep convection, which modifies the boundary layer equilibrium and consequently the updraft mass flux.I have some concerns about the model setup and the derivation of Eq. 1. I suggest a major revision with the following comments.Major comments:1. Uncertainty due to model setupResolution:The effect of explicit versus parameterized deep convection is investigated at a horizontal resolution of 13 km. It is generally believed that the horizontal resolution needed to partially resolve deep convection should be ~1km, the 13km resolution used here is not sufficient to resolve deep convection, so the setup of the S13 experiment would not be recommended. It is unclear how sensitivity is the effect of explicit deep convection to the background horizontal resolution. Will the conclusion be different if a higher horizontal resolution, e.g. 3km, is used?Walls at 30N/S:The aquaplanet simulations has a wall at 30N/S. This setup is likely to effect many spects of the simulations including the ITCZ. It is unclear to me if the conclusion of this study would be different if there is no wall but a global aquaplanet.2, The derivation of Eq. A5, which leads to Eq. 1Eq. A1 is for the top of the BL (the subsidence is w_e) while the equation at L686 is from the balance between radiative cooling and descending is for free troposphere (i.e., the subsidence is not w_e), then, how could these two equations be combined into Eq. A5.Minor comments:
L54: Not sure if "appropriate" is the right word here. Each model center has chosen the model resolution appropriately, according to their needs and computational resources.
L60-70: According to Zhou et al. (2022), the storm-resolving simulation (res ~3km) does not reduce the bias in tropical precipitation characteristics (except for the better representation of strong convection events and tropical cyclones) and is not likely to alleviate the double-ITCZ bias.
Zhou W., L.R. Leung, J. Lu, (2022): Linking large-scale double-ITCZ bias to local-scale drizzling bias in climate models. Journal of Climate 35 (24), 4365-4379.L71: resolving (deep) convectionI suggest moving section 4 (description of the diagnostic framework) to section 2.Citation: https://doi.org/10.5194/wcd-2023-7-RC1 -
AC1: 'Reply on RC1', Hyunju Jung, 19 May 2023
The comment was uploaded in the form of a supplement: https://wcd.copernicus.org/preprints/wcd-2023-7/wcd-2023-7-AC1-supplement.pdf
-
AC1: 'Reply on RC1', Hyunju Jung, 19 May 2023
-
RC2: 'Comment on wcd-2023-7', Anonymous Referee #2, 11 Apr 2023
The novel contribution of this paper is to use the diagnostic framework of Emanuel (2019) to help understand the physical reasons for the differences among a set of 4 experiments using an aquachannel model with prescribed, zonally symmetric sea surface temperature. The 4 experiments differ only in whether or not they use a parameterization of deep convection or one of two parameterizations of shallow convection. The horizontal grid spacing is 13 km, so deep convection is poorly resolved and shallow convection is not really resolved at all. Even so, understanding why the results differ, even if the results are seriously compromised relative to nature, is a step forward, so I am in favor of seeing some version of the paper published with this strong caveat.
The Emanuel framework consists of diagnostic equations for cumulus updraft mass flux, large-scale vertical motion, and a single predictive equation for the mass-weighted vertically integral of the moist static energy. In the original paper, it was used as a tool for very basic understanding of tropical circulations. Here it is being used instead to help diagnose and understand complex simulations, albeit in a simple aquachannel framework with steady, zonally symmetric SSTs.
Of the three equations in the original framework, the current authors use only one. It would be useful if they could explain why they chose only a single diagnostic. The most important criticism I have is that it is not made clear what is being specified and what is being calculated from this framework. I gather from a mediated, anonymous exchange with the authors that the models’ precipitation, surface heat fluxes, radiative cooling, dry static stability, and difference between boundary layer and lower tropospheric moist static energy are being fed into the framework, and precipitation efficiency and updraft mass flux are being diagnosed. Whatever the case, the inputs and output(s) must be clearly stated. The sentence “In our diagnostics, M_u and epsilon_p are not obtained directly from vertical motion but indirectly using other consistent quantities” is far too vague. Perhaps just state that these quantities are diagnosed using (1) and (2) with inputs from the simulations.
One clear difference among the simulations is that the parameterization of deep convection tends to weaken the Hadley circulation. The diagnostic framework does not really help us understand why. Since the output is precipitation efficiency and mass flux and everything else is fed in from the simulations themselves, one would suppose that the focus would be in the predicted quantities. To imply that the Hadley circulations in the simulations with no parameterization of deep convection are stronger because the wind-driven fluxes are stronger seems tautological. When the stronger fluxes are fed into the framework, it dutifully diagnoses a stronger convective mass flux in the ITCZ; not sure what we have learned. I think the authors are up against the age-old problem of inferring causality in a steady system. One might also point out that the specification of SST means that surface energy balance is not enforced; if a slab ocean were coupled it would not be able to sustain the large differences in turbulent heat fluxes observed among the experiments.
One result that is fascinating is the constancy, across experiments, of the precipitation efficiency in the ITCZ region. It would be great if the authors could address this result.
Another improvement that very much help with the understanding of the diagnostic is to plot, either as part of Figure 4 or as a separate figure, the actual terms in (1); namely, the ratio of the surface heat flux to the moist static energy difference, and the ratio of the radiative cooling to the dry static stability.
One other question I have is why the authors chose the diagnostic equation for the cumulus updraft mass flux rather than the one for the large-scale vertical velocity. Is it because the latter is difficult to sample in the simulations? More difficult than sampling rainfall?
A few specific points:
Figure 1: As the authors note, the model does not seem to have settled down into a steady state by the end of the integration. It might be worth it to extend one of the 4 simulations beyond this ending time.
Line 199: “We speculate that extreme rainfalls….”
Line 210-211: It is not necessarily true that the steady state must be equatorially symmetric. There can be spontaneous symmetry breaking.
Equation 2: I would have though that the water vapor concentration that appears here should be evaluated at cloud base rather that taking a vertical average.
Section 5.1.1: I understand the breakdown between wind and delta enthalpy, but why is it important to distinguish sensible from latent fluxes here?
Line 633-634: If radiative cooling is shut off, there can be no latent heating that, over the whole domain, must balance the cooling. The system would shut down.
Citation: https://doi.org/10.5194/wcd-2023-7-RC2 -
AC2: 'Reply on RC2', Hyunju Jung, 19 May 2023
The comment was uploaded in the form of a supplement: https://wcd.copernicus.org/preprints/wcd-2023-7/wcd-2023-7-AC2-supplement.pdf
-
AC2: 'Reply on RC2', Hyunju Jung, 19 May 2023
Hyunju Jung et al.
Hyunju Jung et al.
Viewed
HTML | XML | Total | Supplement | BibTeX | EndNote | |
---|---|---|---|---|---|---|
361 | 145 | 15 | 521 | 41 | 4 | 5 |
- HTML: 361
- PDF: 145
- XML: 15
- Total: 521
- Supplement: 41
- BibTeX: 4
- EndNote: 5
Viewed (geographical distribution)
Country | # | Views | % |
---|
Total: | 0 |
HTML: | 0 |
PDF: | 0 |
XML: | 0 |
- 1