Atmospheric Convection: Known and unknowns
Cathy Hohenegger
Max-Planck-Institut für Meteorologie at least 6 km
Raining
Painting: Layne Johnson, https://laynejohnson.com/product/cumulonimbus/ Height (km) Height
3 mm of rain 34 km/h wind gust
Time (UTC)
Cloud radar located in Lindenberg, picture provided by V. Lehmann (DWD) Under a deep convective cloud…. Why fascination for deep convective clouds?
• Look nice
Painting: Layne Johnson, https://laynejohnson.com/product/cumulonimbus/ Why fascination for deep convective clouds?
• Look nice
• Transport heat, momentum and moisture vertically
Painting: Layne Johnson, https://laynejohnson.com/product/cumulonimbus/ Why fascination for deep convective clouds?
• Look nice
• Transport heat, momentum and moisture vertically
• Produce precipitation
0.1 0.5 1. 2. 3. 4. 5. 10. mm/d Why fascination for deep convective clouds?
Accumulated precipitation, 7 h forecast
• Look nice
• Transport heat, momentum and moisture vertically
• Produce precipitation
• Hard to forecast but can produce severe local damage
Weather forecast from DWD using the ICON model with a grid spacing of 2.2 km Outline
1. Some basics on convection
a. The thermodynamical view
b. The dynamical view
c. The devil messes it up
2. The diurnal cycle of convection over land
3. Food for thoughts Outline
1. Some basics on convection
a. The thermodynamical view
b. The dynamical view
c. The devil messes it up
2. The diurnal cycle of convection over land
3. Food for thoughts Imagine an Earth without convection….
Earth Sun Imagine an Earth without convection…. Imagine an Earth without convection….
• Surface temperature too hot
• Tropopause temperature too cold
• Lapse rate: 11.7 deg/km
Manabe and Stickler (1964) Imagine an Earth without convection….doesn‘t work physically!
• Parcel warmer than environment -> rises
9.8 deg/km • Atmosphere unstable and 6.5 deg/km overturn
• Lapse rate cannot be larger 11.7 deg/km than dry/moist adiabat lapse rate Radiative-convective equilibrium
• Radiation destabilizes the atmosphere
• Convection acts against it and stabilizes it
• Convection wins in the troposphere
Manabe and Stickler (1964) Give large-scale constraints on convection
Temperature (K) 30 • Convection sets the mean temperature structure of the piControl 100 tropical atmosphere RCE 200 • On the scale of the tropics, 300 radiative cooling is compensating Height (hPa) by latent heating from convection 600
1000 220 300 Temperature (K)
Popke et al. (2013) Give large-scale constraints on convection
Jakob, Singh and Jungandreas (2019) Convection arises because of instability caused by radiation Observed dewpoint Observed temperature
• As soon as unstable, 8 12 16 20 24 28 32 convection Dry adiabat Moist adiabat Parcel ascent • Parcel view to check Pressure (hPa) Pressure
Level of free convection Lifting Condensation Level
Temperature (C) Convection arises because of instability caused by radiation Observed dewpoint Observed temperature
• As soon as unstable, 8 12 16 20 24 28 32 convection
• Parcel view to check
Pressure (hPa) Pressure Convective Available Potential Energy • Instability indices
Convective Inibition Energy
Temperature (C) Summary thermodynamical view
• Convection triggered when profile unstable Thermodynamic view Dynamical view
• Column view
• Convection favored in unstable and moist atmospheres
• Large-scale constraint on precipitation amount Pitfall 1: need to know the temperature profile
8 12 16 20 24 28 32 Pressure (hPa) Pressure
Skew-T plot 9 hour before local thunderstorm developed Temperature (C) Pitfall 2: Diurnal cycle
0.3
Precipitation (mm/h) Precipitation 0.2
0 6 12 18 0 Local Time
After Ghada et al. (2019) Outline
1. Some basics on convection
a. The thermodynamical view
b. The dynamical view
c. The devil messes it up
2. The diurnal cycle of convection over land
3. Food for thought Forcing air to go up will also trigger convection
8 12 16 20 24 28 32 Pressure (hPa) Pressure
Temperature (C) Forcing air to go up will also trigger convection: mountains Forcing air to go up will also trigger convection: mesoscale circulation Forcing air to go up will also trigger convection: large-scale circulation Forcing air to go up will also trigger convection: large-scale circulation
Temperature gradient
Pressure gradient
Wind
Wind convergence
Ascent
Back and Bretherton (2009) Summary dynamical view
• Convection triggered by forced vertical ascent Thermodynamic view Dynamical view
• Horizontal view
• Horizontal heterogeneity needed Outline
1. Some basics on convection
a. The thermodynamical view
b. The dynamical view
c. The devil messes it up
2. The diurnal cycle of convection over land
3. Food for thoughts Who is the devil? Who is the devil?
Convection! Thermodynamic effect: Convective clouds moisten their near-by environment
• Convective clouds entrain cold and dry environmental air Detrainment • Leads to evaporation of cloud water Entrainment
• Non-buoyant air is detrained in the environment and moistens it Do undiluted updrafts exist?
• Not over tropical ocean
• But observed in severe convective storms over land
Romps and Kuang (2009) Convection and environmental humidity Moisture-convection feedback
• Convection doesn‘t like dry 90% atmosphere 70% 50% 25% • Convection moistens its nearby environment
• Moisture-convection feedback (see also Grabowsky and Moncrieff 2004)
• Chicken and egg problem….
Derbyshire et al. (2004) Dynamical effect: Convection induces its own circulation Dynamical effect: Convection induces its own circulation
• Latent heating from condensation generates deep heating - + - • Deep heating generates circulation
• Note: in deep convection, latent heating from condensation dominates over other heating sources
• Chicke and egg problem again….
Zonally integrated solution Gill (1980) Observed heating profiles
Land
Ocean
Stachnik et al. (2013) Convection also induces convective outflows in the boundary layer
• Melting of ice and snow and evaporation of precipitation generate non-buoyant air that sinks through the cloud
• Drag from droplet
• Outflow spreads at the surface as a density current, called cold pool Processes happen on different scales! Processes happen on different scales!
Houze (2004) Summary the devil messes it up
• Convection is the devil
• Thermodynamic: convection likes moist atmosphere but convection also moistens its near-by environment
• Dynamic: convection is triggered by circulation but convection generates its own circulation (deep heating + cold pools) Outline
1. Some basics on convection
a. The thermodynamical view
b. The dynamical view
c. The devil messes it up
2. The diurnal cycle of convection over land
3. Food for thoughts The diurnal cycle of convection over land
Cloud top 0.3 Height 0.2
Time (mm/h) Precipitation
0 6 12 18 0 Local Time Idea 1: Changes in cloud morphology during the transition
• Grow wider • Entrain less
Kuang and Bretherton (2009) Del Genio and Wu (2010) Idea 2: Preconditioning: thermodynamical view
Condensates • When the moistening by previous clouds has moistened the atmosphere enough Humidity deviation
0 6 12 18 24 30 36 42 48 Time [h]
Waite and Khouider (2010) Idea 3: Preconditioning: thermodynamical view
Mean transition time (h)
• Mean transition time is 2 h over land
• Preconditioning likely too slow, needs dynamical forcing
Hohenegger and Stevens (2013) Idea 3: Cold pools: dynamical view
• When cold pools form
Khairoutdinov and Randall (2006) Idea 4: Instability: thermodynamical view
• When the lapse rate temperature of the clouds become larger than the environment
Wu, Stevens and Arakawa (2009) So what control the transition?
Shallow Clouds Cold pool Deep Clouds Height
Time Instability idea Cold pool idea So what control the transition?
Shallow Clouds Cold pool Deep Clouds Height
Time Instability idea Cold pool idea Outline
1. Some basics on convection
a. The thermodynamical view
b. The dynamical view
c. The devil messes it up
2. The diurnal cycle of convection over land
3. Food for thoughts Is the timing of convection important ?
• The timing likely depends on Time of max JJA precipitation processes happening on scales O(<10 km)
• Climate models have been run with a wrong timing of convection for decades
Dai (2006) Global simulations at storm-resolving resolution
Cloud rendering as seen from a satellite point of view Stevens et al. (2019) Summary
1. Some basics on convection
Thermodynamic view DynamicalThermodynamic view view Dynamical view
Thermodynamic Dynamical view The devil al view messes it up
2. The diurnal cycle of convection over land
Shallow Clouds Cold pool Deep Clouds Entrainment Preconiditioning Instability Cold Pool