Real-time monitoring and forecast of intraseasonal variability during the 211 African Monsoon R. Roehrig 1, F. Couvreux 1, E. Poan 1, P. Peyrillé 1, J.-P. Lafore 1, O. Ndiaye 2, A. Diongue-Niang 2, F. Favot 1, J.-L. Boichard 3, L. Fleury 3 1 CNRM-GAME, Météo-France/CNRS, Toulouse, France. 2 ANACIM, Dakar, Sénégal. 3 OMP, Toulouse, France
Introduction and Motivation Large progress during the last decade in documenting and understanding the African IntraSeasonal Variability (ISV) 3 main timescales: 25-9 days: link to the MJO (e.g. Matthews 2, Janicot et al. 29) 1-25 days: Quasi-Biweekly Zonal Dipole (QBZD, Mounier et al. 28) Sahelian mode (Janicot et al. 21), Variability of the Saharan Heat Low and link with midlatitudes (Chauvin et al. 21, Roehrig et al. 211) 3-1 days: African Easterly Waves (e.g. Kiladis et al. 26), Synoptic variability of precipitable water (Couvreux et al. 21, Poan et al. 212), Kelvin waves (Mounier et al. 27).
Introduction and Motivation Large progress during the last decade in documenting and understanding the African IntraSeasonal Variability (ISV) 1-25 days: Sahelian mode 3 main timescales: 25-9 days: link to the MJO (e.g. Matthews 2, Janicot et al. 29) 1-25 days: Quasi-Biweekly Zonal Dipole (QBZD, Mounier et al. 28) Sahelian mode (Janicot et al. 21), Variability of the Saharan Heat Low and link with midlatitudes (Chauvin et al. 21, Roehrig et al. 211) 3-1 days: African Easterly Waves (e.g. Kiladis et al. 26), Synoptic variability of precipitable water (Couvreux et al. 21, Poan et al. 212), Kelvin waves (Mounier et al. 27). OLR
Introduction and Motivation Large progress during the last decade in documenting and understanding the African IntraSeasonal Variability (ISV) 1-25 days: Sahelian mode 3 main timescales: Objectives: 25-9 days: link to the MJO (e.g. Matthews 2, Janicot et al. 29) 1-25 days: Quasi-Biweekly Zonal Dipole (QBZD, Mounier et al. 28) Sahelian mode (Janicot et al. 21), Ø Confront the climatology to the real world; Ø Investigate the forecast skill of the monsoon ISV, based on our current knowledge Approach: Ø Address these 2 points through a real-time exercice Variability of the Saharan Heat Low and link with midlatitudes (Chauvin et al. 21, Roehrig et al. 211) 3-1 days: African Easterly Waves (e.g. Kiladis et al. 26), Synoptic variability of precipitable water (Couvreux et al. 21, Poan et al. 212), Kelvin waves (Mounier et al. 27). OLR
Outline 1. Introduction 2. The MISVA project Data and methodology Available products 3. Overview of 211 season. Conclusions
2. MISVA MISVA: Monitoring IntraSeasonal Variability over Africa Collaboration between France & Senegal with the objective to involve forecasters A real-time website, simple but easily and rapidly evolving, according to the encountered needs and ideas Use of websites providing complementary information (broader context) e.g. MJO: Wheeler s website + NCEP Regular briefing & reports (~2/ week) between Toulouse and Dakar. h#p://isv.sedoo.fr
2.a Data and methodology Data: ECMWF 6-hourly analyses and forecasts (+1 days) Daily satellite OLR, with a 2-day delay (NOAA) 3-hourly TRMM precipitation, with a ~6h delay Observations: Senegal Rain gauges, soundings Methodology: Computation of seasonal anomalies, using the mean seasonal cycle derived from ERA-Interim/NOAA OLR/TRMM-3B2 Filtering using forecast data (EC) and zero-padding (as in Wheeler and Weickmann 21) Indexes of ISV modes are computed through a spatial projection of (filtered or not) anomalies on their mean climatological (canonical) structure.
2.a Data and methodology Seasonal anomalies Data: ECMWF 6-hourly analyses and forecasts (+1 days) Daily satellite OLR, with a 2-day delay (NOAA) 3-hourly TRMM precipitation, with a ~6h delay Observations: Senegal Rain gauges, soundings θ85 Canonical structure Methodology: Computation of seasonal anomalies, using the mean seasonal cycle derived from ERA-Interim/NOAA OLR/TRMM-3B2 Filtering using forecast data (EC) and zero-padding (as in Wheeler and Weickmann 21) Indexes of ISV modes are computed through a spatial projection of (filtered or not) anomalies on their mean climatological (canonical) structure. Heat Low index Analyses - 2d Forecasts +1d
2.b Available Products For a start (in 211) we selected just a few products (maps, indices, hovmüller ) Sahelian mode QBZD mode SHL mode Mid-latitude Rossby waves Ventilation Onset Precipitable Water AEWs Precipitation (TRMM & Obs)
3. Overview of 211 ISV: 1-25-day scale HLE 7 June fdb K m s - 1 Shading: θ 85 anomalies Black contours: raw θ 85 Vectors: wind anomalies at 85 hpa Green contours: raw sea level pressure
3. Overview of 211 ISV: 1-25-day scale HLE 7 June fdb Canonical HLE K m s - 1 Shading: θ 85 anomalies Black contours: raw θ 85 Vectors: wind anomalies at 85 hpa Green contours: raw sea level pressure
3. Overview of 211 ISV: 1-25-day scale HLE 7 June fdb Canonical HLE HLW 16 June K m s - 1 Canonical HLW Ø High similarities with canonical events: they make sense in the real world K m s - 1 Shading: θ 85 anomalies Black contours: raw θ 85 Vectors: wind anomalies at 85 hpa Green contours: raw sea level pressure
3. Overview of 211 ISV: 1-25-day scale Heat Low intraseasonal index HLW1 HLW2 HLW3 HLW HLW5 HLE1 HLE2 HLE3 HLE Ø 3 HLE events: 5-9 June, 18-2 July, 1- Sep Ø HLW: 12-18 June, 25 June-1 July, 1-1 Aug,7-1 Sep Ø The season begins with a HLE event, favorable to the monsoon onset (Roehrig et al. 211)
3. Overview of 211 ISV: Onset TRMM PrecipitaVon [1W- 1E] 7.5N Eq. Onset: 13 June TRMM PrecipitaVon [7.5N- ] [- 7.5N]
3. Overview of 211 ISV: Onset TRMM PrecipitaVon [1W- 1E] Heat Low and Onset 7.5N Eq. HLE event = reduced northeastern ventilation The heat low can reinforce Onset: 13 June TRMM PrecipitaVon [7.5N- ] [- 7.5N] HLE1 Favorable to the northward shift of the ITCZ Roehrig et al. (211)
3. Overview of 211 ISV: Onset TRMM PrecipitaVon [1W- 1E] Heat Low and Onset 7.5N Eq. HLE event = reduced northeastern ventilation The heat low can reinforce Onset: 13 June TRMM PrecipitaVon [7.5N- ] [- 7.5N] HLE1 HLE2 Favorable to the northward shift of the ITCZ Roehrig et al. (211) à SHL state provides large scale information for monsoon onset or northward surge.
3. Overview of 211 ISV: 1-1-day scale Synoptic Variability of Precipitable Water (PW) 3N 25N 26 July PW DATE = anomalies 211-7-26 ::. + Wind anomalies at DATE 925 = 211-8-1 hpa ::. 3N 25N 1 Aug 2N 2N 1N 1N 5N 5N W 3W 2W 1W 1E 2E 3E E -8 DATE - = 211-7-28 ::. 8 3N -1-6 -2 2 6 1 28 July 25N W 3W 2W 1W 3N 25N -8-1 -6 3 Aug - 8 DATE = 211-8-3 ::. -2 1E 2E 3E E 2 6 1 2N 2N 1N 1N 5N 5N W 3W 2W 1W 3N 25N -1-8 3 July 1E 2E 3E E - 8 DATE = 211-7-3 ::. -6-2 2 6 1 W 3W 2W 1W -1-8 kg m - 2-6 - -2 1E 2E 3E E 8 2 6 1 m s - 1 2N 1N 5N W 3W 2W 1W -1-8 -6 - -2 1E 2E 3E E 8 2 6 1
3. Overview of 211 ISV: 1-1-day scale Synoptic Variability of Precipitable Water (PW) 3N 25N 3N 25N 2N 2N 1N 1N 5N 5N W 3W 2W 25N -1 1W 1E 2E 3E E -6 28 July W -2 2 6 1 3N 25N 2N 1N 1N 5N 5N 3W 2W 25N -1 1W 1E 2E DATE = 211-7-3 ::. -8 3N 3W - -6 3 July -2 2 3E E -1-6 W 3W -1 kg m- 2-2 1E 2E 1W - -6 2 3E E 8 6 1 t 2W -8 1 1W - DATE = 211-8-3 ::. 3 Aug 8 6 2W -8 2N W t- 3d - ::. 8 DATE = 211-7-28-8 3N Canonical wet PW synopvc event (Poan et al. 212) PW nomalies t 9=25 hpa::. DATE =a211-7-26 ::. + Wind anomalies a DATE 211-8-1 26 July 1 Aug 1E -2 2E 2 3E E 8 6 1 m s- 1 t+3d! 2N 1N 5N! W 3W 2W -8 1W - 1E 2E 3E E 8 Ø High similarities between with the canonical event: In structure, in dynamics: The canonical event makes sense -1-6 -2 2 6 1 kg m- 2 m s- 1
3. Overview of 211 ISV: 1-1-day scale Synoptic Variability of Precipitable Water (PW) Ø Same characterisvcs as the canonical mode: frequency, propagavon Ø Differences between the eastern and the western Sahel structures (as in Poan et al 212) PW anomalies [12N- 2N] kg m - 2
3. Overview of 211 ISV: 1-1-day scale Synoptic Variability of Precipitable Water (PW) Ø Same characterisvcs as the canonical mode: frequency, propagavon Ø Differences between the eastern and the western Sahel structures (as in Poan et al 212) PW anomalies [12N- 2N] Precipitable Water index and Rainfall: over Senegal PW anomalies (kg m - 2 ) PW anomalies Rain gauge precipitavon Precip (mm day - 1 ) June July August Sept Ø High correlation (.57) between precipitation and PW over Senegal, especially after the monsoon onset (.63). Ø High potential of the PW variable kg m - 2
Conclusions - Perspectives Conclusions: 1-25 days: Monitoring of heat low variability provides some large-scale information on the monsoon system. The relationship with the onset seems to work this year The monitoring of the QBZD and Sahelian modes is less obvious (not shown). Synoptic scales: PW has a large potential: high predictability, strong relationship with precipitation. Perspectives: Quantitative evaluation of: Diagnostics (e.g., filtering effects) Skills of ECMWF forecasts for different scales (PW, Heat Low ) This exercise continue this year with new diagnostics and with higher involvement of Senegalese forecasters
Beginning of the 212 season TRMM PrecipitaVon [7.5N- ] [- 7.5N] False start SHL intraseasonal index TRMM PrecipitaVon [1W- 1E]
Beginning of the 212 season