Becky Venton, a chartered meteorologist and advisor to Resurgence on weather and climate services , writes on a groundbreaking, creative collaboration to make weather forecasts for Nairobi and Dar es Salaam more locally accurate and relevant.
Designing for the needs of key city users
In support of the DARAJA Project (Developing Risk Awareness through Joint Action), I visited the Kenya Meteorological Department (KMD) in Nairobi and the Tanzania Meteorological Authority (TMA) in Dar es Salaam. The aim of my visit was to work with the forecasters of both organisations to develop more locally accurate forecasts for both Nairobi and Dar es Salaam and to tailor these to better meet the needs of the DARAJA stakeholders, namely the residents of informal settlements and the organisations that support them.
Through previous project activities including Information Ecosystem Mapping, household surveys and other workshops, a number of specific user needs had already been identified. These included: presenting the weather information in a simpler format; using less technical language; and including details of the potential impacts of the weather. The residents of the informal settlements in both cities also wanted details relating to the amount, duration and likelihood of rainfall, with a specific request for the weather forecasts to be more localised.
The quest for localised forecast information
The request for more localised forecast information is in no way surprising, especially within cities the size of Dar es Salaam (1,590km2) or Nairobi (696km2) but it does present some challenges. In my home country, the UK, we are very lucky to have a supercomputer that is one of the most powerful in the world dedicated to weather and climate. Alongside advances in science and underpinning technology, increased computing power has enabled the Met Office to improve the accuracy of its forecasts and the lead times for predicting high impact weather events. This has in part been achieved through increasingly high resolution modelling, where the side of each grid square covering the UK has been reduced to less than 1.5km in length, and by improvements made to the models themselves so that they more accurately represent the dynamics of the atmosphere.
Since 2011, under the World Meteorological Organization’s (WMO) Voluntary Cooperation Programme (VCP), the Met Office has been running a 4.4km Africa model which KMD, TMA and other National Meteorological and Hydrological Services (NMHS) in Africa can access and use. Under a sister project of DARAJA within the Weather and Climate Information Services (WISER) East Africa programme, new science has been introduced to this model. Through the HIGH impact Weather lAke sYstem HIGHWAY) Project, the model is now ‘convective permitting’ and as a result, it can better represent severe thunderstorms. However, even with this development, whilst forecasters may be able to accurately predict that there may be rain showers in an area, it remains difficult to accurately predict the exact location where rain will fall other than when convective clouds are more ‘organised’ (such as around a tropical storm)
Finding other pathways to local precision
So what can be done? A number of National Meteorological and Hydrological Services (NMHS) have the capability to run their own ‘regional’ models that take output from the GPCs to produce a higher resolution output. At the time of my visit to Tanzania, TMA were in the final stages of testing their 5km resolution model and in Kenya, KMD hope to have a similar model running in the near future. These are exciting developments and I am pleased to know that TMA’s model is now operational. However, at the time of the workshop, it was not certain that the data from a local model would be available during the lifetime of the DARAJA project. I therefore wanted the forecasters to consider other ways that they could make the forecast more locally accurate.
I chose my words carefully – ‘more locally accurate’, rather than ‘downscaling’. Knowing that for many, the term ‘downscaling’ can be synonymous with running higher resolution NWP models. I was challenging the forecasters to ‘think outside of the box’, that is, to think beyond the ‘grid box’ of computer-based numerical weather prediction and to consider other ways that a forecast can be made more locally precise.
More locally accurate and more locally relevant
During the workshop a number of great ideas emerged, including developing local knowledge about how models perform in certain locations and comparing temperatures in the project area with those measured at the nearest ‘official’ NMHS maintained site. Other suggestions were: to gather and utilise indigenous or traditional knowledge; to analyse past weather to understand any local weather effects; to consider climatological zones; and also to harness knowledge and experience relating to local climatology from other NMHS staff across their organisations. Armed with new insights as to how to make the forecast ‘more locally accurate’, the next challenge was to make the forecast ‘more locally relevant’. This was to be achieved through language, impact statements, weather icons and advisory messages made possible by excellent work already undertaken by the DARAJA stakeholders to produce reference guides.
At the end of each workshop, the new and enhanced daily and weekly forecasts for Nairobi and Dar es Salaam were presented to a wider group of DARAJA stakeholders, who were delighted with the developments and had suggestions for further improvement. Both KMD and TMA took on board all of the comments and worked hard to update the forecast templates in the light of user feedback.
The new forecast templates for Dar es Salaam are now being prepared for dissemination by TMA. In the case of KMD, there is positive initial feedback from the end users about the new format of the city forecast which provides weather information for two separate zones within Nairobi County and also uses new weather icons. Ultimately, these improvements will lead to greater understanding and use of weather and climate information in Nairobi and Dar es Salaam.