Availability and Quality of Weather and Climate Data
Data can be available at all times, but its the quality and reliability are the paramount issues. Weather and climate data must be made available online for the accessibility of end-users like pilots and other flight operators, including the general public for easy planning and decisions making. The information must be a real representation of the atmosphere so that it can fit, and contribute to what is expected from the meteorological community. If a weather observer should refuse to provide the actual weather due to ignorance or deliberate, then it tends to affect the whole meteorological organization. This kind of report will not agree, and comply with the surrounding, accurately observed from other stations, and will not support the ‘value chain’ that the meteorological body is trying to arrive at, in order to boost the image of meteorology. Models that take their ‘bases-run’ from these dubiously observed weather information will always have a sort of challenges in their output. This is because it will not represent the real atmosphere, and in event, the forecast will not be reliable. In this regard, automatic weather observing stations have been provided to assist in weather observations, but the maintenance at most stations for continuous operation is quite challenging.
Author: (Senior Meteorologist), Ghana Meteorological Agency, P. O. Box 87, Legon - Accra, Ghana.
Weather products, like models and satellite imageries have been made available for Meteorologists to use in making daily and extended forecasts for decisions to be taken by the population. Images from satellites and other technical documents, like models are made available on the internet. Lack of access to this information should not be an excuse this time in the 21st Century. Visual weather observations made at Meteorological Stations are now available on the internet, for this reason individuals can monitor the weather, even privately for their personal comfort and use.
During the 2018 HAJJ pilgrimage from Tamale in Ghana to Mecca, an incoming pilot, in his ‘post flight report’ complained that the TAF and METAR for the Tamale Meteorological station must be made available online for easy flight planning. The compliant was accepted by the responsible officials and the issue was quickly resolved. Other flight operators like the Air Traffic Controller and Flight Dispatchers could now access Tamale weather online.
Satellite imageries are quite accurate weather documents that describe the weather and climate of the Earth, if only they are well interpreted. There are different types of satellite imageries, and they are near-real-time images, not prognostics, they take selfies of the Earth and had the date and time inscribed on them for the sake of analysis. Subsequent watches or successive views of these satellite imageries can lead to the prediction of the weather for a place. These lots of weather products in the system will need a choice to describe the atmosphere.
EUMETSAT has made available numerous satellite products, with RGB composites, on the internet such that any mobile phone or computer that is connected to the internet can access satellite imagery. To mention a few, there is Dust RGB satellite imagery which monitors dust particles in the atmosphere (Figure 5). For this matter areas on the surface or in the atmosphere with hazy conditions could be seen and monitored. Different colours are seen on the same pictures frame which has their respective meanings.
The previously available satellite imageries were just infrared and visible imageries. Visible channels are not useful in the night because they function on the reflection of sunlight. The infrared channels are useful throughout both day and night. It has made the choice of meteorological products a tactical one, especially when a forecast is demanded. Satellite products previously consist of black and white colours, where white indicate clouds and black on the imageries signifies clear skies. Even with that, certain techniques have to be deployed in other to distinguish between cloud types and other atmospheric constituents like dust.
Latest innovations include the representation of the synop codes plotted on maps (Figure 6). These are perfect ways of digesting weather information in the meteorological office. Satellite observations are validated with the ground truth information from the various meteorological offices. Weather plots are a representation of weather elements that is happening at the station (at the time of observation) and what has already happened in the past few hours. A concise description of the station which includes various cloud types, the transparency of the atmosphere, air temperature and pressure, as well as humidity, and the precise location are indicated on a map for easy analysis to be made.
In numerical weather prediction, equations have been built into high-speed running computers that factor the rotation of the Earth into account before producing the output as forecast. Distances, usually termed as grids are highly respected in this kind of production because the Earth continues to rotate and revolve at the same constant speeds.
Winds charts from the Weather Re-analysis and Forecast model by the Ghana Meteorological Agency seem not to be on grid points and looks quite suspicious, and a little bit different from those that are produced by other models like the ECMWF, NOAA, UKMO and Meteo France. Looking at the product critically (Figure 1), adjacent wind barbs do not lie on consistent longitude and latitude as the others do, and wind barbs in the southern hemisphere still look like those in the northern hemisphere. Analyzing them seem quite challenging. Sometimes it becomes very difficult to determine in which direction the wind is blowing especially when the winds are light . When there is a situation like this, the choice of product to use in making a forecast would have to go to the most consistence product, the one that everybody would agree to its production.
The seriousness in it is that, weather forecasters would lose confidence in using the product and would not even like to evaluate the products from inconsistent models. Whether the product is doing well or not would not be a concern to weather chart analysts. Resources would be wasted in the production while other models continue to be famous in weather chart production.
Weather and climate data become quality when the producer knows exactly why he or she is giving out the information. Rainfall report in Meteorology, for example, is made such that an insight of the cloudy conditions of the day can be inferred from the main synoptic hour reports. Rainfall is reported every six hours, and has been given codes from 1 to 4, in multiples of 6 hours to make up the 24 hours in the day. The idea is to add subsequent rainfall amounts and report them at 0600 UTC, in the morning as a multiple of 4 of 6 hours rainfall for a station. 1200 UTC will be 1 of 6 hours of rain after 0600 UTC , 1800 UTC will be 2 of 6 hours of rain totals . For this reason, if it should rain before 1200 UTC it must reflect in the 1200, 1800, 0000 and 0600 UTC weather reports respectively for someone to know that the place had been very cloudy during a certain period of the day.On the 6th May, 2018 Accra Met Office failed to report the totals of rain at 1800 UTC  which occurred before 1200 UTC. These are things that define product quality, the observer might not know the reason for reporting the rain again at 1800 UTC, so it makes the report quite misleading in terms of reviewing and when making weather forecasts for the station. So many parameters are considered when running models and if the right information is not fed into the system then the out-puts could be a bit challenging.
Just to compare winds models from GMet (Figure 1) and Meteo France (Figure 3); Meteo France, like the other models indicate a dot at the point of observation and circles around the dots where winds are calm but GMet model leaves it blank, making the product so different from the others. All these dots lie on consistent longitudes and latitudes producing perfect squares with every four adjacent points across the chart (Figure 3).
Figure 2. and figure 3. have their winds originating from uniform longitudes and latitudes on the wind chart which form the known grid points. The spatial distribution of the wind barbs in Figure 1. does not follow the trend for which perfect grids could be manifested; the winds not originate on particular longitudes and latitudes when observed critically when making analyses.
MATERIALS AND METHODS:
Methods and materials include the retrieval of weather charts from various models like the ECMWF, NOAA, METEO FRANCE, and from the GMet WRF models, and analyzing them. Organizing daily and weekly weather chart discussions with Meteorologists in the GMet Forecast Office. Training newly employed Meteorologists in Ghana, and also university graduates on attachments, and National service personnel.
These training includes the analysis of surface pressure charts, wind charts at standard levels and the interpretation of satellite imagery of various kinds; RGB composites, Infra-Red, Water Vapour, and Visible imageries, and to compare with the ground observations made at various synoptic stations in Ghana and around the Africa continent.
RESULTS AND DISCUSSIONS:
It has been realized that analyzing winds from the GMet models does not look easy as the other models do because of the non-grids of the wind barbs, and also identification of wind direction when they are less than 3 knots. Sometimes determination of the wind direction in areas of ridges, troughs, and vortices become difficult to decide due to uncertainties.
Dialogue with producers of this model suggested that the problem could be due to the computer program used in producing the model and efforts are on the way to resolve these problems.
It would be very necessary to have meteorological reports checked thoroughly, including the coding of synop from the Meteorological stations. Also, the renovation of Meteorological offices, and the capacity of personnel as well as training and motivation of worker, so that they can be dedicated to the work and give out their best to make weather and climate data a real representative of what it is supposed to be.
A visit to most of the stations in Ghana shows how some stations are under-staffed, some instruments broken down, including the destruction of enclosure fences which puts the few instruments left vulnerable to threats by inhabitants in the community. Some stations need to be relocated due to encroachment which is seriously affecting weather elements.
There were instances where a storm or a squall could affect a station with very strong winds but the instrument, the anemometer would not be able to indicate the actual speed of the wind, compelling the observer not to record such a phenomenon as a squall.
It would also be very good if model users in the institutions could be tasked to investigate and validate products from different models that are available, and document them for consistency. This would help meteorologists to be able to make choices of products in relation to the seasons.
Figure 1: Winds barbs chart at 925 hPa level from GMet WRF model for 6th May, 2018 at 1200 UTC .
Figure 2: Wind vectors chart at the 200 hPa by NOAA for 13th May, 2018 at 0000 UTC .
Figure 3: Winds barbs chart at 925 hPa by Meteo France for 5th January, 2018 at 1200 UTC .
Figure 4: Post flight information form from a pilot requesting for Meteorological data to be published online.
Figure 5: Dust RGB satellite imagery from EUMETSAT on 18th June, 2018 at 1515 UTC indicating dust storms over Mali, Niger, and Saudi Arabia (Magenta), and rain storms over the Gulf of Guinea .
Figure 6: Synop plots for Ghana and neighboring countries on 18th June, 2018 at 1200 UTC showing surface weather observations .
NOAA (National Oceanic and Atmospheric Administration),
MESA (Monitoring for Environment and Security for Africa),
PUMA (Permission to Use EUMETSAT in Africa), GCAA (Ghana Civil Aviation Authority),
GMet (Ghana Meteorological Agency),
ECMWF (European Center for Medium range Weather Forecast),
UKMO (United Kingdom Meteorological Office).
- https://www.ogimet.com/display_synopsc2.php?lang=en&estado=Ghan&tipo=ALL&ord=REV&nil=SI&fmt=txt&ano=2018&mes=05&day=06&hora=12&anof=2018&mesf=05&dayf=06&horaf=12&send=send (synop for Ghana on 06/05/2018 at 1200 UTC)
- https://www.ogimet.com/display_synopsc2.php?lang=en&estado=Ghan&tipo=ALL&ord=REV&nil=SI&fmt=txt&ano=2018&mes=05&day=06&hora=18&anof=2018&mesf=05&dayf=06&horaf=18&send=send ((synop for Ghana on 06/05/2018 at 1800 UTC)
- GMet Forecasting System
- NOAA African Desk
- MESA, PUMA (synergie) Meteo France.
(SYNOP from 65472, Accra (Ghana) | 05-36N | 000-10W | 68 m
201805061200 AAXX 06124 65472 11465 72704 10291 20242 40130 69901 70262 8117/
333 58014 81814 87460 =) 
(SYNOPS from 65472, Accra (Ghana) | 05-36N | 000-10W | 68 m
201805061800 AAXX 06184 65472 32465 72305 10290 20241 40119 8483/
333 10310 58019 82610 83816 87360= 
SYNOPS from 65442, Kumasi (Ghana) | 06-43N | 001-36W | 287 m
201805061800 AAXX 06184 65442 11470 62304 10265 20230 40133 69922 70362 84230
333 10278 58035 84814 86360= 
SYNOPS from 65445, Sefwi Bekwai (Ghana) | 06-12N | 002-20W | 171 m
201805061800 AAXX 06184 65445 11464 72702 10260 20246 40126 69902 70162 8153/
333 10285 58040 81611 87362=)