Department of Meteorology Internal Variability in the...
1
Ros Cornforth Brian Hoskins Chris Thorncroft Department of Meteorology Questions To what extent is rainfall predictable? From Figs. 1 and 2 make the hypothesis: • AEWs are a synoptic system capable of organising convection Then use the idealised approach to explore: What impacts AEW evolution, growth/decay, structure and location? 1. Impacts on AEJ-AEW Evolutions 2.Impacts on AEJ-AEW Structures 3. Geographical dependencies Time-series of (a) AEJ max wind speed in different expts and (b) 3D moist AEJ wind max and AEW max EKE Internal Variability in the West African Monsoon and 8-Day Rainy Sequences The Challenge Predictions of rainfall in sub-Saharan Africa show high uncertainty. Short-range forecasts (up to 2 days) are needed for the public and aviation, and medium-range forecasts (10-30 days) are key for agriculture, hydrology and health information. AEJ-AEW System is a crucially important component of West African monsoon (WAM) rainfall variability. Many convective systems move with and through AEWs, with periods in which convection intensifies and vice versa (Laing et al, QJ, 2009). In Fig. 2, convection is modulated on an AEW-like length-scale. But systematic errors plague the forecast skill in the region and AEWs decay far too quickly in forecasts (Agusti-Panareda et al, ECMWF Newsletter, 2008). Forecasting is still nowcasting. Atmospheric processes are incorrectly represented. Urgent need for underpinning theoretical framework to amend NWP models. Acknowledgements • The project was funded by NERC and is affiliated to the Walker Institute for Climate System Research: www.walker-institute.ac.uk . Summary 1. Credible dry and moist simulations of the African Easterly Jet (AEJ) and African Easterly Waves (AEWs) that grow on it, show that these core dynamical features of the West African Monsoon are strongly inter-dependent throughout their life cycle, hence “the AEJ-AEW system”. 2. The inter-dependencies impact the evolutions, growth/decay, structures and locations of the moist AEJ- AEW system, which are consistent with observations. 3. These dependencies establish an internal 8-10 day variability between the model rainfall and periods of growth and decay in the AEWs - consistent with intra- seasonal observations of rainy day sequences (Fig. 1). Cornforth et al (2009). Q. J. R, Metorol. Soc., 135, 841-913 Email: r.j. cornforth @reading.ac. uk The Approach: Use Idealised Modelling 1. Readingʼs Intermediate Global Circulation Model: idealised yet retains important interactions. Integrated first with dry and then with moist physics to diagnose the impact of moist processes. 2. Prescribed zonally symmetric surface temperature and moisture profiles establish the AEJ through meridional contrasts in dry and moist convection. AEJ is forced not prescribed (Fig. 3) 3. AEWs initialised by breaking zonal symmetry and integrating model with zonal wavenumber 13 symmetry, after randomly perturbing surface pressure coefficients at t=0. Fig. 3 (a) Moist processes contribute to faster developments of moist AEJ and AEWs than in dry life cycle. (b) Slow oscillations in moist AEJ time series. Periods of deceleration in AEJ coincide with max AEW growth. AEW Growth/Decay and Rainfall • Speculation of an external forcing causing observed rainy day sequences (Sultan et al, J Clim, 2003) may well be explained instead by this internal forcing in the moist AEJ-AEW system. • Significant result from the moist life cycle, providing an alternative explanation for some of the observed intraseasonal variability of the precipitation over West Africa. • It has implications for the evaluation of weather and climate prediction models for West Africa. 1. Moist AEWs have intermittent periods of decay and growth. Growth is preceded by increased mean rainfall. 2. Complex interactions between AEJ, AEWs, moist convection and upper levels, leads to an internal variability on a time-scale of 8-10 days EKE of moist AEW Model rainfall at max AEW lat 3D moist AEJ (include waves) 2D moist AEJ 2D dry AEJ (a) (b) 3D moist AEJ EKE of moist AEW Zonal mean cross-sections averaged over days 15-35 of the dry and moist life cycles (d) (d) Zonal wind - dry AEJ Zonal wind - moist AEJ Ertel PV (black) & Theta Heat low ccn EKE of dry AEW ITCZ + Heat low ccn Moist Ertel PV (black) & Theta EKE of moist AEW 3xweaker cf moist AEW Polewards low-level amplitudes in both dry/moist Diabatic heating in moist life cycle increases meridional PV gradient at mid-levels. Moist processes -> contraction of scales. Positive PV anomaly EKE of moist AEW EKE of dry AEW Latitude of dry AEJ Latitude of moist AEJ (a) (b) Latitudinal displacements of the dry and moist AEJs correlate with the growth/decay of AEWs. In (a), the dry jet moves equatorwards whilst in (b) the moist jet moves polewards. This is consistent with observations. Fig.1: Daily rainfall time series (mm) from June to September 1968 averaged over the grid points from 10W to 10E and 12.5N and 15N (white) and in black the corresponding index filtered above 60 days. Fig.2: Satellite image of water vapour. Coldest cloud tops indicated in red (see scale in top RH corner) Observed 8-10 day rainy sequences
Transcript of Department of Meteorology Internal Variability in the...
Ros Cornforth Brian Hoskins Chris Thorncroft
Department of Meteorology
QuestionsTo what extent is rainfall predictable?From Figs. 1 and 2 make the hypothesis:
• AEWs are a synoptic system capable oforganising convection
Then use the idealised approach to explore:What impacts AEW evolution, growth/decay,structure and location?
1. Impacts on AEJ-AEW Evolutions
2. Impacts on AEJ-AEW Structures
3.Geographical dependencies
Time-series of (a) AEJ max wind speed in different expts and (b) 3D moist AEJ wind max and AEW max EKE
Internal Variability in the West AfricanMonsoon and 8-Day Rainy Sequences
The ChallengePredictions of rainfall in sub-Saharan Africa show high uncertainty.Short-range forecasts (up to 2 days) are needed for the public andaviation, and medium-range forecasts (10-30 days) are key foragriculture, hydrology and health information.
AEJ-AEW System is a crucially important component of WestAfrican monsoon (WAM) rainfall variability. Many convectivesystems move with and through AEWs, with periods in whichconvection intensifies and vice versa (Laing et al, QJ, 2009).
In Fig. 2, convection is modulated on an AEW-like length-scale.
But systematic errors plague the forecast skill in the region andAEWs decay far too quickly in forecasts (Agusti-Panareda et al,ECMWF Newsletter, 2008). Forecasting is still nowcasting.
Atmospheric processes are incorrectly represented. Urgent need forunderpinning theoretical framework to amend NWP models.
Acknowledgements• The project was funded by NERC and is affiliated to the Walker Institute
for Climate System Research: www.walker-institute.ac.uk.
Moist processes impactthese dependencies as well
Summary1. Credible dry and moist simulations of the African Easterly
Jet (AEJ) and African Easterly Waves (AEWs) that grow onit, show that these core dynamical features of the WestAfrican Monsoon are strongly inter-dependent throughouttheir life cycle, hence “the AEJ-AEW system”.
2. The inter-dependencies impact the evolutions,growth/decay, structures and locations of the moist AEJ-AEW system, which are consistent with observations.
3. These dependencies establish an internal 8-10 dayvariability between the model rainfall and periods ofgrowth and decay in the AEWs - consistent with intra-seasonal observations of rainy day sequences (Fig. 1).
Cornforth et al (2009). Q. J. R, Metorol. Soc., 135, 841-913Email: [email protected]
The Approach: Use Idealised Modelling1. Readingʼs Intermediate Global Circulation Model: idealised yet retains
important interactions. Integrated first with dry and then with moistphysics to diagnose the impact of moist processes.
2. Prescribed zonally symmetric surface temperature and moisture profilesestablish the AEJ through meridional contrasts in dry and moist convection.AEJ is forced not prescribed (Fig. 3)
3. AEWs initialised by breaking zonal symmetry and integrating model withzonal wavenumber 13 symmetry, after randomly perturbing surfacepressure coefficients at t=0.
Fig. 3
(a) Moist processes contribute to faster developments of moist AEJ andAEWs than in dry life cycle. (b) Slow oscillations in moist AEJ time series.Periods of deceleration in AEJ coincide with max AEW growth.
AEW Growth/Decay and Rainfall
• Speculation of an external forcing causing observed rainy day sequences (Sultan etal, J Clim, 2003) may well be explained instead by this internal forcing in the moistAEJ-AEW system.
• Significant result from the moist life cycle, providing an alternative explanation forsome of the observed intraseasonal variability of the precipitation over West Africa.
• It has implications for the evaluation of weather and climate prediction models forWest Africa.
1. Moist AEWs haveintermittent periods ofdecay and growth. Growthis preceded by increasedmean rainfall.
2. Complex interactionsbetween AEJ, AEWs, moistconvection and upper levels,leads to an internalvariability on a time-scale of8-10 days
EKE of moist AEW
Model rainfall at max AEW lat
3D moist AEJ (include waves)
2D moist AEJ
2D dry AEJ
(a) (b)3D moist AEJ
EKE of moist AEW
Zonal mean cross-sections averaged over days 15-35 of the dry and moist life cycles
(d) (d)
Zonal wind - dry AEJ Zonal wind - moist AEJ
Ertel PV (black) & Theta
Heat low ccn
EKE of dry AEW
ITCZ + Heat low ccn
Moist Ertel PV (black) & Theta EKE of moist AEW3xweaker cf moist AEWPolewards low-levelamplitudes in both dry/moist
Diabatic heating in moist life cycle increases meridional PV gradient at mid-levels. Moist processes -> contraction of scales.
Positive PV anomaly
EKE of moist AEWEKE of dry AEW
Latitude of dry AEJ Latitude of moist AEJ(a) (b)
Latitudinal displacements of the dry and moist AEJs correlate with the growth/decay of AEWs.In (a), the dry jet moves equatorwards whilst in (b) the moist jet moves polewards. This isconsistent with observations.
Fig.1: Daily rainfall time series (mm)from June to September 1968averaged over the grid points from10W to 10E and 12.5N and 15N (white)and in black the corresponding indexfiltered above 60 days.
Fig.2: Satellite image of water vapour.Coldest cloud tops indicated in red(see scale in top RH corner)
Observed 8-10 day rainy sequences
