Notes

All the images shown on these webpages have been generated from the National Centers for Environmental Prediction (NCEP) Global Forecast Systems (GFS) model. A thorough description of this model can be found here. We use analysis and forecast products with a resolution of 0.5x0.5 degrees every six hours. At present, the data we use has a forecast period of 84 hours (3.5 days).

Our plots are generated every six hours at 00UTC, 06UTC, 12UTC and 18UTC (most plots are available approximately 15-20mins after these times) and while this process occurs some images/sequences may be slightly odd or just missing - just have a coffee and check again! The plots are delayed by approximately 12 hours due to time required for NCEP to assimilate the data and run the forecasts, transmitting the forecasts and local processing of the data. We use Unidata's GEMPAK package to make all our plots (more information about GEMPAK is available here). We make use of code written by Dave Vollaro (SUNY Albany) to compute Potential Vorticity from the pressure level GFS data (analysis times only - see note in section c(ii)) and code written by Anantha Aiyyer (SUNY Albany) to calculate streamfunction. The objective trough/jet plotting code was written by Gareth Berry (SUNY Albany) after development in collaboration with Tim Hewson (UK Met Office) and Chris Thorncroft (SUNY Albany).

The trough/jet diagnostics are what we believe to be the main 'value added' product available on these pages. We use a completely objective method to identify African Easterly Waves (AEWs) and jet axes based on the 700hPa vector wind field. Essentially, trough lines are defined as the point at which the advection of the curvature vorticity is equal to zero, in regions where the curvature vorticity is positive. Similarly, jet axes are defined as the point at which the shear vorticity is equal to zero, with masks added to isolate wind speed maxima. Further masks are added to show trough lines and jet axis in certain flow regimes that are specific to our interest. We only plot trough axes in regions where there is easterly flow, to differentiate between AEWs and e.g. mid-latitude troughs. We only show jet axes in region where the magnitude of the wind is in excess of 8m/s to remove regions of weak flow. A paper documenting these diagnostics and their application to the AEWs of July, August and September 2004 (authored by Gareth Berry, Chris Thorncroft and Tim Hewson) has been accepted to MWR (subject to minor revisions) and an early copy of it is available here. A more rigorous description of these diagnostics (including the relevant equations and nice examples) can be found here. These diagnostics are overlaid on all our horizontal maps.

PV is shown with the trough lines in order to illustrate the multi-scale aspects of AEWs, as discussed in our recent case study paper (Berry and Thorncroft (2005), MWR), which can be accessed here. The trough lines tend to emphasize the wave-like nature of AEWs that is often discussed in the literature and it may be possible to interpret the horizontal tilts of the trough lines relative to the jet axes as evidence of barotropic growth. The PV field may help to re-enforce the notion of dynamical waves (via a wave-like perturbation of the zonally orientated PV strip near 10N), but more commonly show the intense diabatically generated PV maxima that are embedded within the AEW (and via the invertability principle may make up a substantial contribution to the AEW flow anomaly), that from our experience appear to be directly associated with tropical cyclogenesis in the ocean basins. The 700hPa wind vectors are shown in blue to aid interpretation of the objective troughs.

On a technical note, we use the 315K theta surface because it is located near 700hPa in the vicinity of 10-15N (the AEW 'action' zone). This material surface bows down over the Sahara desert, and reaches as low as 925hPa near 25N in the height of the summer (see our 925hPa theta plots for confirmation). An important point to note when examining the analyses and forecast is that the methods used to compute PV for the forecast and analysis times are slightly different. Although there may be some small discrepancies in the amplitude of an individual PV maxima, the main difference is that for the forecast periods we see very strong and widespread PV maxima that form over the desert regions during night hours (due to strong low-level radiation cooling that produces large static stability values near the Earth's surface). However, these problems occur away from the main areas of interest, so we don't believe it to limit the use of our plots.

These fields are shown primarily in order to allow the tracking of westward moving vorticity centres that we expect to be moving along the low-level baroclinic zone as part of an AEW (see our recent paper here for an example of such a feature). We suggest these features are indicative of baroclinic energy exchanges associated with AEWs. The latitudinal location of the baroclinic zone will also give some indication of the progress of the West African monsoon.

These fields are part of our selection intended to look for a relationship between AEWs and convection. We expect that if a vorticity centre is propagating along the low-level baroclinic zone (as part of AEW) it will effect the theta-e distribution at low-level and may impact convection (also discussed in our paper). The 400hPa vertical velocity field (with only values less than -8 pa/s (i.e. Ascent) contoured) indicates the location of strong active convection.

These diagnostics are generated in an attempt to determine where regions that favour long-lived deep convection are located. We know from past studies that we tend to see such systems in regions with pronounced low-level wind shear (top left panel), instability (indicated by the k-index (coloured) in the top left panel) and dynamical forcing (we have chosen 925hPa moisture flux convergence). Three model soundings are shown (10N 10E, 10N 0W and 10N 10E) in order to look for mid-level dry layers (which generates downdrafts and promotes cold pool propagation in a sheared environment) and to examine the vertical shear profile throughout the depth of the troposphere.

Three cross-sections at 10W 0W and 10E are shown. The top figures show mixing ratio to show the structure of the monsoon layer and zonal wind in superimposed to show the location and intensity of the African easterly jet (AEJ) and the tropical easterly jet (TEJ), two significant features of the summertime atmospheric circulation in this region. For ease of viewing, the wind values are multiplied by minus one so that easterlies are contoured with a solid line. The lower figures show PV (coloured) so that we may get some insight into the vertical structure of the PV maxima that are embedded in the AEWs. Potential temperature contours are also shown to illustrate the structure of the low-level baroclinic zone and the Saharan heat low.

The Atlantic plots are a continuation of the 315K PV plots that are shown for North Africa (see the description of those above), although without a background satellite image. They are generated in the hope of allowing us to determine the ultimate fate of AEWs. From experience we have noted that the vast majority of AEWs can be followed as either a distinct PV anomaly or trough line (sometimes both if the AEW is vigorous) into at least the Caribbean (i.e. 60W) or the mid-latitudes. We have noted that on many occasions, our trough diagnostics have strong resemblances to the Tropical Prediction Center (TPC) hand analyzed surface map for the Atlantic ocean and Caribbean sea. To see this, compare our latest analysis with the appropriate analysis time (wide area: 00UTC 06UTC 12UTC 18UTC, SW Atlantic: 00UTC 06UTC 12UTC 18UTC) from the TPC.

Similar to the diagnostic described above, these are a continuation of the 315K PV plots that are shown for North Africa. Instead of a basin wide view, these show a more zoomed in area, centred on the Carribean and Gulf of Mexico. The diagnostics are overlaid on GOES IR imagery, which we recieve via CIMSS, University of Wisconsin.

In the top image, our trough/jet diagnostics are overlaid on the CIMSS (University of Wisconsin)/Hurricane Research Division (NOAA) Saharan Air Layer (SAL) diagnostics. These diagnostics indicate the presence of dry, dust laden Saharan Air in the lower troposphere (below 500hPa) over oceanic regions as yellow/red shading. The presence of this airmass is thought to detrimental to the formation of tropical cyclones. The image used is copyright of UW-CIMSS/NOAA HRD and is used here with permission the original images, with full details on how they are constructed are available from the UW-CIMSS SAL webpages.

The lower panels show diagnostics from the GFS model, which we hope further help to delineate the SAL from other air masses and to demonstrate the value of the Satellite derived product. The lower left panel shows relative humidity, averaged between 850 and 500hPa, in order to show dry air over the continent and give an appreciation of how close the GFS analysis is to Satellite observations. The lower right shows potential vorticity averaged between 850 and 500hPa in order to give an indication of the potential source of a particular air mass. We expect that air from the Saharan boundary will have very low PV due to the almost neutral stability as a result of dry convection - hence it may be possible to differentiate the dry SAL from other dry air masses (e.g. dry subtropical air) and its relationship with AEWs by using both the satellite derived products with the GFS output.