1
Françoise Guichard (1) , Laurent Kergoat (2) , O. Bock (3) , F. Timouk (2) and E. Mougin (2) Variability of the daytime sahelian boundary layer sampled at Agoufou via tethered balloon and kite flights in August 2006 1)CONTEXT A combination of tethered balloon and kite flights has been carried out over the Malian site of Agoufou (located at 1.5W-15.3N, within the AMMA northern site); boundary layer (BL) data were collected almost daily from the 12 to the 31 August 2006. Pressure, temperature, humidity, wind speed and direction together with CO2 concentration were measured either entirely over the whole depth of the BL at a high spatial (~20m) resolution (vertical profiles up to 2km AGL whenever possible, with a ~3h time sampling), or in the mixed layer when the wind was too strong for balloon flights (~100-400 m AGL). This dataset will allow extending the analysis of CO2 surface emission achieved with nightime boundary-layer data acquired in 2004 and 2005; they provided estimation of CO2 respiration fluxes over the Agoufou Savannah site (Kergoat et al. 2005). With 2006 data, daytime CO2 fluxes will be estimated for golden days, via convective boundary layer budgets – taking advantage of collocated surface measurements. The temporal coverage and spatial resolution of this dataset allows to study the characteristics and the day-to-day variability of this scarcely documented sahelian boundary layer, during the core of the monsoon season, and to assess its representation in models. Figure 1 (on the right side) provides information on the monsoon season time-height series of atmospheric dynamics and thermodynamics at the observation site, - ECMWF analysis. The lower panel of Fig.1 indicates that the period of measurement started by a series of rainless days, that were following a more rainy phase. It was interrupted by the passage of a single squall line the 18 Aug (10 days after the previous rain event), before the occurrence of a more rainy period after the 21 Aug (periods Tdry and Tmoist). Some disctint features characterize these two periods in the analysed atmospheric local field as well. 2) HIGHLIGHTS Fig. 2 illustrates the often strong diurnal BL dynamics (RH & wind), as well as significant day-to-day-changes. * The sampled convective boundary layer typically grows from a few hundreds of metres in the mid-morning up to 1.2 km or more by the early afternoon (Fig. 3) – measurement amazingly consistent with Tombouktou soundings on a day-by-day basis except when one site too close in time from disturbed situations (convection). * The BL wind speed is often stronger in the morning, with strong wind shear at the top of the BL where the wind speed increases significantly (Fig. 2). Strong BL wind speeds persist all day however on a 2 days (21 & 28 Aug, Fig.2), which happened to be the most contrasted in terms of mixed-layer properties as shown in 3). Finally, the measurement period took place during a transition from an 3) BOUNDARY LAYER CHARACTERISTICS Mixed-layer properties are summarized in Figs. 4, 5 & 6. * distinct diurnal fluctuations emerge (e.g.; CO2, T, RH, |V|) in Fig. 4 * Overall the period Tdry is warmer & dryer than Tmoist (consistent with station data Fig. 5) * thetae (Tdry) is higher too, even though rv(Tdry) is slightly less. It leads to a negative correlation linking thetae and RH (Fig. 6) when considering the whole 20-day period, i.e. a distinctive signature * Not intuitively, [CO2] is lower during Tdry, suggesting stronger vertical mass exchanges during Tmoist (Fig. 4) which need to be quantified * The mixed layer properties of the 2 windiest days are almost marking out the range of fluctuations of thetae and [CO2], indicative of the importance and complexity of some advective processes. * The 21 Aug is characterized by the highest rv, but is among the dryest in 4) BOUNDARY LAYER HUMIDITY, CLOUDS AND WATER * lower BL RH suggests the possibility of lower cloud amounts. Inferences on the cloud cover from surface radiative fluxes point to this statement, they also show significant compensations between surface shortwave and longwave fluxes so that the variations of the net surface radiative flux between Tmoist and Tdry are minimized during this phase of the monsoon. * the differences of IRT histograms between Tmoist and Tdry are consistent with Tmoist being more cloudy that Tdry, this is also obvious from IRT time series (Fig. 7) – there is much more scatter during Tmoist (spread of cloud top heights). Precipitable water (PW) data from the GPS station of Gao (0W,16.25N) show a progressive decrease of the atmospheric water vapour amount, briefly and abruptly disturbed by the passage of the 18 Aug squall line (sharp positive fluctuation, on the order of the rainfall amount, and ~ 25% of PW), then a rapid increase within less than 2 days, starting the 21 Aug, after which PW remains overall higher until the 31 Aug (Fig. 8). This feature is well seen by the ECMWF analysis, which further points to a large scale nature of the phenomenon,with a progressive drying followed by a strong moistening accompanied by a shift to the North of the maximum of PW (this pattern still stands out for wider longitude averages). IRT histograms over a wider 10°x8° area also point to large fluctuations of the cloud cover from the 18 to the 23 Aug (Fig. 8). The atmospheric moisture deficit could have play some role in the build up and maintenance of high BL thetae, as well as on the type of convective events that occurred. Understanding the full chain of causalities which accounts for this functioning remains an object of focus for future studies. 18 Aug 24 Aug 28 Aug 19 Aug 23 Aug 21 Aug 20 Aug mixing ratio rv rel humidity RH late afternoon, ahead of a squall line following day windy fair weather strong monsoon flow disturbed afternoon following an early morning MCS fair weathe r windy all day, MCS outflow imprint 26 Aug Figure 2 : boundary layer data obtained for a selection of days Figure 4 : boundary layer data obtained for a selection of days orange Tdry days, blue Tmoist days except 21/8 green 28/8 grey... (color code as in Fig 2.) Figure 3 : diurnal dynamics of convective BL heights Agoufou balloon data ECMWF Analysis Tombouktou soundings Heights of boundary layers Figure 5 Figure 6 : 2-m relative humidity (obs) 12 Aug 21 Aug 30 Aug (%) (1): CNRM/GAME (CNRS & Meteo-France), (2) CESBIO, (3) IPSL/SA & LAREG/IGN Acknowledgments Based on a French initiative, AMMA was built by an international scientific group and is currently funded by a large number of agencies, especially from France, UK, US and Africa. It has been the beneficiary of a major financial contribution from the European Community's Sixth Framework Research Program. Detailed information on scientific coordination and funding is available on the AMMA International web site http://www.amma-international.org. We thank Yakouba Traore who actively participated to the collection of the data during the field campaign, Hamma the people from Agoufou for their help. M. N. Bouin(LAREG/IGN) and E. Doerflinger (LDL, CNRS) are acknowledged for their involvement with the GPS instruments, We thank K. Ramage from IPSL for his help with the IRT data. Figure 7 : histograms and times series of IRT for the 2 periods Tdry and Tmoist dry period (Tdry) moist period (Tmoist) , min and max values IRT : 50km average (280 pixels) Figure 8 : time series of PW from GPS at Gao and IRT ~ 50km wide domain around Gao, histograms of IRT over 5W-5E,12N-15N and time- latitude serie of PW, surface pressure and wind vectors from ECMWF analysis 18 19 20 21 22 23 Aug
