| |
Research
Climate change in Africa
Future lower-tropospheric moisture flux anomalies in East Africa
23 July 2009
This page provides supplementary information for the
manuscript Shongwe et al. (2009). The material presented below supplement the
discussion on "Features relevant for the future atmospheric water budget" in East Africa ( i.e. their Section 4).
Anomalous moisture flux convergence is found over East Africa. Westward moisture transport from the tropical Indian Ocean is likely to strengthen. This includes a combination of moisture transported downstream of the south-easterly monsoonal flow and by the north-easterly monsoon winds. In a few cases, a reversal of the climatological westward moisture flux at tropical (10 °N - 10 °S) longitudes west of ~30 °E is projected. This manifests as anomalous eastward moisture flux from the east Atlantic Ocean across the Congo basin. These forms of moisture flow are projected to work together to enhance precipitation in East Africa. Our results show that in East Africa, an increase in tropospheric water vapor content could enhance precipitation in the face of a possible slow down of the tropical circulation (Vecchi and Soden, 2007). The stronger increase in East Africa is a result of local effects.
Background information
As the climate warms, observational, theoretical and modelling studies agree with regard to an increase in precipitable water. Increases in lower-tropospheric water vapour at a rate ~7%/° C are to be expected from the Clausius-Clapeyron relationship. However, atmospheric dynamical effects result in spatial inhomogeneities in the global warming-induced tropical rainfall increases. For instance, projections for rainfall in East Africa are higher than the zonal mean. We show here that this is in part caused by anomalous moisture flux convergence over East Africa.
An analysis of lower-tropospheric (below 500mb level) horizontal moisture flux anomalies across East Africa during the major rainy seasons (OND and MAM) has been carried out. The relevant ingredients required to compute the moisture flux anomalies Q' are:
- V; the horizontal wind vector with compontents u and v,
- q; the atmospheric specific humidity, and
- g; the gravitational acceleration.
The vertical integral is taken from the 1000hPa (pb) to the 500hPa (pt) pressure level. The angular bracket denote the time mean for future ( f), and present ( p; defined as the 1961--2000 period).
To characterise the interannual variability, dry (wet) seasons in the 1961-2000 period have been identified for each model whenever the amplitude of the first standardised principal component score of simulated rainfall is less (more) than one standard deviation. Wet minus dry composites are constructed by an equation similar to the one above using wet minus dry seasons instead of future minus present. The resulting moisture flux anomalies are displayed in the first columns of the Figures below. The composites are then compared with future (2081-2100) moisture flux anomalies (second columns).
Anomalies of moisture flux convergence are shaded in the figures below.
Moisture flux anomalies during the short-rains (OND)
Moisture flux anomalies calculated for the OND season are displayed below. We show here results from five CGCMs. Flux anomalies projected by the other five models are shown for the MAM season. Anomalous moisture flux convergence over the western Indian Ocean and the inland area including East Africa (dashed rectangle) is evident in the composites (first column). By virtue of being sandwiched between rich moisture sources (i.e. the Indian Ocean, the east Atlantic Ocean and Congo Basin), East Africa benefits in various ways. Firstly from strengthening of the climatological westward flux from the Indian Ocean. This occurs particularly if the southeasterly flux (from south of the equator) converges with the northeasterly flux (from the north) or if either stream decelerates over East Africa. Secondly, from a reversal of the climatological westward flux at Eastern Hemisphere tropical longitudes west of ~30 °E, and thirdly by a convergence of these anomalous fluxes. In the model future (2081-2100) simulations, anomalous westward moisture flux by easterly trade winds south of the equator is found in e.g. the CGCM3.1, GFDL-CM2.0 and ECHAM5/MPI models (Figs. b,f and h), and in other models not shown. The CSIRO-Mk3.0 model (Fig. j) represents projections in which anomalous moisture transport from the east Atlantic Ocean across the Congo Basin occurs, which is however not very common in the ensemble used here. It is noteworthy that although the pattern of anomalous moisture transport is not similar in the models shown here, the net effect in almost all models favours an increase in East Africa precipitation in the future climate through anomalous convergence. These results are consistent with the projected mean precipitation changes presented in Sec. 3a.
Although there are similarities between the CGCMs future projections of short-rains change and the composites of interannual variability (especially in the IPSL and MIROC models), there are also large differences (e.g., in the CNRM, GFDL and MRI models). We can therefore not expect to explain all factors that cause more rainfall in East Africa by mechanisms that are known from interannual variability.
October-December (OND) lower-atmospheric horizontal moisture flux anomalies ( kg m-1 s-1). For each model (name given in the title), the first column displays composites of the wettest minus driest OND seasons during the 1961-2000 period. The second column shows the 2081-2100 moisture flux anomalies with respect to the 1961-2000 period. The grey shading shows areas of anomalous moisture flux convergence. The dashed rectangle shows the boundary of East Africa as defined in this study.
Moisture flux anomalies during the long-rains (MAM)
During the long-rains (MAM), the pattern of anomalous moisture transport is broadly similar to that found during OND. The three patterns of anomalous moisture flux favourable for enhanced precipitation in East Africa emerge also during MAM. In the wet minus dry composites, anomalous eastward transport across the Congo Basin is found in CNRM-CM3 and GFDL-CM2.1 models (Figs. a,c). The future anomalous pattern in these models differ from the wet-dry composites in the current climate. Anomalous moisture influx from the west Indian Ocean is found instead. Despite these differences, either pattern leads to an increase in East Africa precipitation. In the other models (IPSL-CM4, MRI-CGCM2.3.2 and MIROC3.2), anomalous westward moisture transport across the west Indian Ocean feature prominently. In some cases, moisture transported by the northeasterly monsoons (north of the equator) and easterly influx downstream of the south-east trades (south of the equator) is found (Figs. f,h,j). From this one can conclude that East Africa is a beneficiary of the global warming-related increase in tropospheric precipitable water through favourable circulation anomalies during transition seasons: Firstly, the easterly flow across the tropical Indian Ocean is projected to increase its contribution to the moisture transported into East Africa. This involves a combination of moisture transported in the south-east trades and north-easterly monsoon current. Secondly, the anomalous westerly flux (i.e. a reversal of the climatological easterly flux) from the east Atlantic across the Congo Basin projected by a few models should provide an additional ingredient to enhanced precipitation in East Africa. The pattern of climate change and interannual variability do not show a close resemblance during MAM.
Same as above (i.e. OND season but for the long-rains (MAM) season.
| |
|
