Description

In this finished study we have investigated the Saharan heat low (SHL) dynamics with regard to the interactions of the SHL with the monsoon flow to the South and the maritime boundary layer to the West. A key tool were numerical weather simulations using the meso-scale COSMO-model (V3.19). We used COSMO to perform an operational forecast in support of the GERBILS (GERB Intercomparison of Longwave and Shortwave radiation) campaign which took place in West Africa from 18 June to 31 June 2007. GERBILS aimed to investigate the radiative feedback of dust on the thermodynamics in the region of the southern edge of the Sahara. Sixth scientific flights with the FAAM BAe 146 research aircraft were conducted between Niamey, Niger and Nouakchott, Mauritania.

model domain for COSMO GERBILS forecasts

figure 1: model domain for COSMO GERBILS forecasts, the colour shaded section shows model orography; the grey shaded large domain indicates surface temperature.

The COSMO forecasts were computed on the HP XC6000 supercomputer at the Steinbuch Centre for Computing (SCC) at the Karlsruhe Institute of Technology (RZ-news 2007/03 (german)). We used the ECMWF analyses and forecasts as initial and boundary data. The COSMO runs were initialised at 00 UTC and 12 UTC and computed for 72h. The model domain covered major parts of Mauritania and Mali and thus the standard flight track along 18°N (figure 1). The horizontal resolution was 0.0625° (7km) and 35 levels were used in the vertical.

Aircraft and dropsonde data were used to validate the COSMO model in a desert environment.

figure 2: The aircraft (black line) and dropsonde (red lines) data was interpolated to produce a section of potential temperature (top) and humidity mixing ratio (bottom). The following figures show these sections at the right, sections from corresponding model data at the left. Keep in mind, that the interpolation routine causes artefacts where no data is available, e.g. in particular at lower-levels at the western edge of the section. See thesis for model validation on individual profiles in the monsoon, Saharan and coastal region.

The character of the measurements determines an emphasis on the heat and moisture budget. The comparison showed that COSMO represents the vertical structure of the atmosphere accurately. It locates the different features of the West African monsoon, such as the monsoon layer, the convective internal boundary layer, the Saharan residual layer, and the Saharan Air Layer (SAL) at correct heights. The horizontal position of these features differs slightly from measurements. The height of the top of the SAL is remarkably well predicted. This leads amongst other things to good forecasts of mid-level altocumulus clouds at the top of the SAL. The potential temperature in the model seems to be generally around 1K too high and humidity mixing ratio around 1 g/kg too low. As COSMO does not include dust it struggles with representing fine layers (observable in meteorological quantities as well) which are highly linked to dust loading. The validation encouraged us to use COSMO for further investigations of the temporal and spatial evolution in the south-western Saharan region.

The analysis of the temporal and spatial evolution of the baroclinic zone in the west of the Saharan heat low using COSMO data, showed a complex mesoscale meteorological system with a distinct diurnal cycle at the Mauritanian coast, which we call the "Atlantic Inflow". This system consists of the sea breeze front at the western Atlantic coast of West Africa, the baroclinic zone marking the transition towards the SABL/monsoon layer, and the perturbation of the mid-level troposphere through an induced gravity wave and a distinct frontal circulation.


figure 3a: schematic sketch of the Atlantic Inflow and characteristic features	figure 3b: cross section of vertical velocity (blue: ascent, red: descent), virtual potential temperature and horizontal wind at 18°N

The horizontal extent of the Atlantic Inflow front becomes evident by a strong horizontal temperature gradient and an increase in wind speed combined with a change in wind direction to westerlies behind the front.

figure 4: magnitude of virtual potential temperature gradient (shaded and black contours, with a contour interval of 0.05 K/km), horizontal wind vectors (black arrows, m/s), and the 288 K dewpoint temperature isotherm at 950hPa.

At 15 UTC a stationary front becomes established at the coast. At 18 UTC the front starts moving inland in the low levels and it reaches the Tagant mountains (more than 400km inland) at around 2 UTC. The Atlantic Inflow shows deepest inland penetration south of 20°N and north of the Inter Tropical Discontinuity (ITD).

The terrain in western Mauritania is favourable for the inland penetration. South of 20°N a wide coastal plain extends from the coast at 16°W to the foot of the Tagant mountains at 12°W, which form the first orographic barrier east of the coast and reaches 400-600 m above mean sea level. The sea breeze front is a result of the strong temperature gradient between the ocean and land surface. The cold Canary current and upwelling deep waters lead to a cold ocean surface (22°C) and a rather cool air layer at the Mauritanian coast (see figure 1). In contrast the strong daytime insolation heats the land surface to more than 45°C. Thus a strong temperature and density gradient is evident along the Mauritanian coast. However as an analysis of the heat and moisture budgets in the COSMO forecasts revealed, during the day turbulent mixing due to dry convection over land hinders the inland movement of the cooler air. The cooling due to horizontal advection is balanced by warming due to turbulent diffusion. In the late afternoon horizontal advection dominates and the low-level front starts moving inland, reaching the Tagant after midnight.

The analysis of the heat and moisture budgets also showed that through its advection of cool and moist maritime air the Atlantic Inflow has an important impact on the regional heat and moisture budgets.

For a more detailed discussion we refer to our contribution to the QJRMS Special Issue on AMMA WP 2.1:

C. M. Grams, S. C. Jones, J. H. Marsham, D. J. Parker, J. M. Haywood, V. Heuveline, The Atlantic Inflow to the Saharan heat low: observations and modelling. Quart. J. Roy. Meteor. Soc., 136(s1), 125-140, doi:10.1002/qj.429 (http://www3.interscience.wiley.com/journal/122394603/abstract).