Rapid urbanisation and the urban heat island, a case for Models in the Cloud
Lagos, Nigeria is one of the fastest growing cities in the world. The cities’ population has doubled since 2000 to 13 million and is set to double again by 2030 (UN). This rate of expansion will have undoubtedly lead to social and infrastructure risks to the city.
One of the main impacts from rapid urbanisation is the local warming known as the urban heat island effect. Anthropogenic changes to the Earth’s surface, for example grey construction materials, absorb and trap incoming solar radiation during the day and release it slowly at night. Whilst this warming is ubiquitous worldwide, little consideration has been given to cities outside the Northern hemisphere, in part due to paucity of meteorological observations.
Working with the interdisciplinary Ensemble team, new techniques are being used to classify urban land use using historical Landsat satellite imagery over Lagos. These are used as input to the Weather Research and Forecasting model (WRF), a community based numerical weather prediction scheme that is used both for operational forecasting and atmospheric research. This complex model is being implemented into the cloud-computing infrastructure by the Models in the Cloud initiative at Lancaster University. This will enable pressing environmental issues to be modelled without the need for traditional high-performance computing clusters, and reduce technical, model installation barriers to environmental scientists, myself included.
Initial simulations for February 2016, a period of notable hot weather, have shown a dominant downwind warming caused by a south-westerly wind flow over the city. Pending evaluation, the mean night-time warming within the city is up to 3oC. Captivatingly, a large rural warming caused by heat advection from the city over 1oC extends tens of kilometres downwind. Following evaluation, the model will be run for historical land use scenarios to see how the urban heat island pattern has changed.
This work will be followed by air quality modelling for the Lagos, using the chemistry extension to the WRF model, to see if the dominant south-westerly flow is likely to cause exacerbated health risks from pollutants advected to rural areas.
Author: Richard Bassett