Formation of extended near-surface stratus clouds in West Africa

In recent years, members of the working group "Atmospheric Dynamics" have discovered, described and explained a phenomenon that has been unknown in its spatial extent: nocturnal near-surface stratus clouds in southern West Africa. These clouds affect a region of at least 800000 km² (i.e. more than twice the area of Germany) and are most extensive and frequent between June and September.

The clouds were first described by Schrage et al (2007). Based on radiosonde ascents in Parakou (Benin, roughly 9°N, 2°E) during the monsoon in 2002, the authors found that cloudy nights were characterized by stronger turbulence and convergence in the large-scale wind field. Knippertz et al. (2011) describe the connection with the formation of a nocturnal near-surface jet stream and the absence of these clouds in certain satellite products as well as in global circulation models. With data from the AMMA field campaign in 2006 from Nangatchori (Benin, about 100 km northwest of Parakou), Schrage and Fink (2012) were able to demonstrate that a strong nocturnal jet stream combined with not too high static stability near the surface leads to a sudden transition from a laminar to a turbulent flow after midnight. This turbulent situation occurred in about 2/3 of the investigated nights. Schrage and Fink (2012) assumed that the turbulent eddies vertically mix the near-surface moisture up to the condensation level. This could be proven by means of complex simulations with the Weather Research and Forecasting (WRF) Model (Schuster et al. 2013). This work showed that during monsoon nights additional stratus-forming processes (e.g. cold air advection from the tropical Atlantic) as well as processes that prevent its formation (e.g. inversions, advection of dry air) occur (see Fig. 1). In the end, small differences decide whether the nights remain clear or whether a cloud deck forms. Schuster et al. (2013) also show that the stratus in the model extends inland from the ocean after sunset. This is confirmed by the first stratus climatology by van der Linden et al. (2015), which uses different satellite instruments over a 5-year period: From June to September and starting from the coastal region at 6 pm, the cloud cover gradually extends until it reaches maximum in the late morning (9-10 local time) over a region of at least 800000 km² (i.e. more than twice the area of Germany) (Fig. 2).

Fig. 1: Nocturnal jet stream at the coast of West Africa, associated advection (ADV) of temperature and specific humidity, cooling [moistening] of the cloud region by turbulent fluxes of heat (H) [specific humidity (E)]. Once formed, cooling by long-wave radiation stabilizes the stratus. The units are in Kelvin and in g/kg. From Schuster et al. (2013).

Fig. 2: Stratus frequency in %, June-September 2007-2011. In the morning at 10 UTC (about local time) the stratus region extends from the coast to 10°N and influences an area of about 800000 km². Very often stratus occurs over the Ivory Coast, over Nigeria higher cloud layers often block the view of the stratus (shading). From van der Linden et al (2015).

Why is this phenomenon so important?

Because the stratus does not dissipate until around noon, it significantly reduces solar irradiation and soil temperature in southern West Africa. In WRF model configurations that include the Stratus, showers and thunderstorms occur later in the day and less intense. In the study by Hannak et al (2017), it was again demonstrated that global climate models have great difficulty in simulating the monsoonal stratus. The reasons are complex and partly lie in model errors regarding the reproduction of the afternoon convective boundary layer: Some models start the night too dry in the evening.

What are the main research questions?

Besides the role of the stratus for the overall monsoon in West Africa, the following two questions are still unanswered: (a) What role does human activity via extreme air pollution and heavy deforestation in West Africa play for the stratus cloud cover and the resulting occasional thunderstorm rain? (b) Which processes are responsible for the late dissolution of the low cloud cover during intensive short-wave irradiation? Answers to these questions will be provided by the EU-project DACCIWA (2013-2018), which is coordinated by our working group and in the course of which a large-scale international field campaign took place during the monsoon 2016 (Flamant et al. 2018).

What role do warm humid air masses play in the development of winter storms?

By analyzing a modified form of the pressure tendency equation, members of the working group "Atmospheric Dynamics" have been able to estimate the contribution of latent heat release to the explosive deepening of four severe European winter storms in the last 15 years. For the storms Xynthia (2010) and Lothar (1999) the contribution was up to 60%, while for Martin (1999) and Kyrill (2007) it was almost half as much. This is a relevant contribution to the question of the strength of winter storms in a warmer and wetter world.

Falling surface pressure, which often indicates the approach of a low pressure system, means that the mass of air in the vertical column above the barometer is decreasing. If the barometer were to be carried along in the core of a migrating low-pressure system and the low-pressure system were to intensify, i.e. the barometer pressure were to drop, this would mean that mass would be evacuated from the atmospheric column above the migrating low-pressure system. How can this happen? One possibility is that the air column warms up - quite intuitively, because warm air is lighter than heavy air. However, this does not always have to lead to pressure fall in the atmosphere, as is schematically explained in Fig. 1. As a result of heating, the upper edge of the column can expand upwards, the mass remains the same, but is distributed over a vertically thicker layer. In reality, this will happen to some extent, but the heating will also cause mass to be evacuated laterally.

What causes heating or cooling of the air column?

This is shown schematically in Fig. 2. In most cases, the column is heated by warm air advection, whereas it is cooled due to rising motions. Due to the dominance of latent heat release in the clouds as a result of condensation as well as cloud and precipitation formation, diabatic processes cause a heating of the box and therefore usually contribute to the pressure fall. The pressure tendency equation estimates the net effect of the three temperature-changing processes and relates it to the change in height of the box (in practice via the change in geopotential height in 100 hPa). However, Fink et al. (2012). could only diagnostically estimate the diabatic contribution in the pressure tendency equation as a residual by using the observed pressure drop in the target box.

What was the result?

Apparently the diabatic contribution was particularly pronounced during the explosive development of the European storms Xynthia in February 2010 and Lothar in December 1999. Up to 60% of the pressure fall was due to the diabatic term, whereas in the case of Martin (December 1999) and Kyrill (January 2007) it was only about 30% (Fink et al. 2012). An application to more than 50 cyclones shows a broad spectrum of baroclinic and diabatic contributions, with storms that stay long to the south of the upper-level jet - and therefore in warmer and more humid air - showing more diabatic amounts (Pirret et al. 2017).

Fig. 1: A warming of the air column leads partly to a fall in surface pressure and partly to an increase of the upper edge of the air column.

Fig. 2: Within a target box into which the storm moves (dashed arrow), the pressure tendency equation analyzes, among other things, the net effect of horizontal temperature advection (TADV), of vertical movement (VMT) and of warming by condensation, estimated by the term evaporation (E) minus precipitation P. The target box extends from the ground (psfc) to a height of 100 hPa (about 14 km). From Fink et al. 2012.

Why is the investigation of the diabatic-driven pressure fall so important?

Climate projections project an average weakening of the temperature contrast between subtropical and sub-polar latitudes over the Atlantic. This results in a lower number of winter storms over the North Atlantic and Europe in the future (Zappa et al. 2013). However, under these less frequent favourable conditions of a strong temperature contrast, there is the potential for a few very heavy winter storms due to a concurrently increased moisture content. These, in turn, cause the most damage in Europe. Therefore, it is very important to understand the contribution of warm humid air masses in the present and future climate to the explosive deepening of East Atlantic storms.

What are the most important research questions?

The used method of the pressure tendency equation is not yet fully developed and would have to be applied to a larger sample of observed winter storms. Technical questions include, among other things, what changes occur when the diabatic term is explicitly present, what influence the shape and size of the box in Fig. 2 has, and what errors result from the vertical and horizontal discretization. In addition to the question of which significance precipitation has for the development of low pressure systems, the pressure tendency equation also opens up the possibility of investigating what role the stratosphere can play by lowering the upper edge of the air column (Pirret et al. 2017) - in Fig. 2, this means that the air flows out to the side of the atmospheric column under study. This process may have been significant during the Braer storm in January 1993, which reached a core pressure of 914 hPa near Iceland and was the deepest low in recent decades (Odell et al. 2013).

Why are sporadic dry season rains predictable in West Africa?

Members of the working group "Atmospheric Dynamics" have for the first time provided deep insights into the atmospheric process chain, which is associated with heavy rainfall in West Africa during the dry season. Since the trigger is a low pressure system penetrating into the tropics from the mid-latitudes of the Atlantic-European region, there is the potential to warn the population a few days earlier.

In January 2004, more than 50 mm were recorded within 48 hours in Benin (7-11°N) and its neighbouring countries in West Africa (Knippertz and Fink 2008) - on average, no more than 5 mm falls in the north of the region during the dry season in January. During the dry season, the rains caused the forest savannah to turn green (see Fig. 1). In the case study of January 2004, Knippertz and Fink (2008) show that an extratropical upper trough (in principle a low-pressure system at an altitude of 5-9 km) penetrates the tropical belt from Northwest Africa. This is associated with increasing upper winds and warm air advection on the east side of the trough. Often a slanted band of clouds, a so-called "tropical plume", is formed here, the global climatology of which we published in Fröhlich et al. (2013). Upper warm air advection reduces the surface pressure deep into the tropics, i.e. in latitudes around 10°N. Even though the pressure fall is only a few hectopascals, it is sufficient to shift the equatorial trough (also called heat low), which is located at about 7-8°N, a few 100 km to the north within the flat pressure field. Behind it, humid air masses from the tropical Atlantic penetrate into West Africa and the strong afternoon warming triggers the severe tropical storms in it (cf. Fig. 2). In such a situation, the climatological high over Libya disappears and a low pressure zone extends from the Mediterranean to the Gulf of Guinea. These findings were directly implemented into practical weather forecasting in a book published in 2017 entitled "Meteorology of Tropical West Africa: The Forecasters' Handbook" with African meteorologists (Conforth et al. 2017).

Fig. 1: After heavy rainfall in the last decade of January 2004, the forest savannah in central Benin has turned green - normally one would see withered and partially burnt grass. Source: A. Fink.

Fig. 2: Scheme of the processes involved in the intense winter rains. Involved components are the tropical trough over North Africa ("upper trough"), the warm air advection, the northward displaced heat low and equatorial low pressure trough ("ITD, Intertropical discontinuity", compare dotted green line as climatological position) and the zone of advection of humid air (blue arrows) and tropical rain (blue shaded zone). There is also rainfall north of the Sahara. From Davis et al (2013).

Why is the study of this phenomenon so important?

Firstly, winter rains have positive and negative socio-economic effects. These range from the rotting of cotton stored in the open to the easier ploughing of the soil before the rainy season begins. On the other hand, Knippertz and Fink (2009) and Davis et al. (2013) show that a trough situation over North Africa is already apparent one week before the event, resulting in a high probabilistic predictability of up to one week for tropical convection. In other words: An early warning of the population seems possible. Furthermore, the decoded process chain can serve as a blueprint for extratropically influenced tropical rain at other seasons and in other regions. In this context, a study is currently underway that suggests a similar mechanism for the occurrence of extreme tropical storms in the pre-rainy season in the northern Congo region in February. In addition, an increasing vertical wind change in the lowest 4 km due to the extratropical influence plays a role here.

What are the most important research questions?

Knippertz and Fink (2008) hypothesize that part of the pressure fall over the Sahara and Sahel is due to a locally and temporally enhanced greenhouse effect resulting from the tropical plume and increased water vapor content. The atmospheric column is warmed through positive radiation balance and thus, the pressure falls. Does the increased occurrence of such interactions through this mechanism explain the disproportionately strong warming of the Eastern Sahel in late winter and spring observed in recent years? And what does interaction look like outside the winter half-year, resulting in an improved predictability of summer monsoon rains in West Africa? The latter is a goal of the project W2W (C2).

The mystery of "giant" dust particles

Dust from the Sahara (bottom) is frequently blown towards America (top). The long-range transport of coarse dust has been underestimated so far. (Image: NASA Earth Observatory).

Buoys in the middle of the tropical Atlantic, thousands of kilometres away from the west coast of Africa, repeatedly detect “gigantic” dust particles. With a diameter of up to half a millimetre, they are actually too big and heavy to stay in the air that long. For this reason, such particles have not yet been taken into account in climate simulations, although they may have important effects on cloud formation, radiation and the oceanic carbon cycle. Possible explanations are uplift in thunderstorms or tropical cyclones, the effects of electrical fields caused by the friction of dust particles against each other and turbulent mixing. However, a quantitative estimate of these effects based on the current state of knowledge shows that it only seems possible to transport particles of the sizes found that far across the ocean by combining very unlikely events. So there remain a number of extremely exciting questions that must now be addressed in order to reduce the uncertainty in dust models.

Further information: van der Does et al. 2018