Welcome to my Geospatial portfolio. I use this space to display my capabilities and skill in GIS and map design. Feel free to browse through past projects I have worked on and my contribution to the yearly 30 day map challenge.

BREAKER

There are those who follow maps, and those who make them.

This map was the final output of a series of maps and geo-computational analysis charts assessing the possibilities of expanding the agricultural crop output in the East of England and reducing river pollution levels, all whilst considering projected climate change effects on seasonal precipitation and temperature levels.
Using univariate regression on UKCP climate data, historic and current Land Cover Maps and River pollutant data, I was able to project statistically, that there is a high probability that arable cropland will be able to expand by 19% by 2040 compared to 2007. This was then visualised by highlighting (yellow on map) 19% of the available grassland, bare grassland, and low-productive rough ploughing fields (most likely land types to become suitable for arable farming) that were most distant from river networks.

The task was to aid a large research project called STOP (Science and Technology in childhood Obesity Policy), led by Imperial College Business School. A member of their team reached out to my previous employer to source as much open-source data for locations of secondary schools, fast food outlets and greenspaces/parks/sports centres in specific London boroughs.

For context of the above visualisation, I had chosen the borough of Tower Hamlets as this met a set criterion, being:
• High concentration of low-income households
• Low spatial distribution of greenspaces (compared to suburban Greater London)
• Inner city borough (expected high concentration of fast-food outlets)
• Large number of schools for small area to meet demand from dense population

The map was created in QGIS, using multiple open datasets including the road network from Ordnance Survey, boundaries, greenspaces, and schools’ points data from London datastore, and fast-food location data from Public Health England. Most of the data is overlaid on each other and an underlying heatmap was made to visualise the spatial density of fast-food outlets. 400m buffers were made around each school to act as a catchment area to assess the average school to fast-food ratio and the school to greenspace ratio.

Description of data visualisation explained above in map.

For context, this map was created as a challenge given to me as part of the interview process. The challenge was to use any free, open-source data provided by a specific geospatial portal and create an informative map outlining levels of risk relevant to the insurance sector. I chose earthquake shaking data, as this is one of the most popular reasons for large residential/commercial property damage insurance claims in this region, thus, emphasising the importance and relevance for a visualisation of this risk. As the task highlighted to keep the result simple and it being only a visualisation, rather than an analysis output, I took advantage of using bold symbology to make the areas susceptible to higher risk stand out and opted to use a non-satellite/hybrid base map to keep the bodies of water less striking to the eye.

In early 2021, I was appointed the task to process and extract CMIP global climate change data to produce country-level statistics and NetCDF files for the Bank of England. The statistics were published as part of the assessment of physical risk variables under climate change scenarios. The intended purpose was to achieve a collection of benchmark statistics for each variable, so that member organisations can align their model outputs within range of the benchmark. I was one of three providers of physical risk data for the 2021 CBES.

Summaries of my work can be found on the final two pages of the supplementary guidance. View Document

The above figures visualise the main steps I took in this task:
• Aggregate all NetCDF data from CMIP portals (Figure 1)
• Use python (GDAL & CDO) to read data and align/flip extents (Figure 2)
• Use country boundary shapefiles to mask global data (Canada in F2 & Figure 3)
• Create python script to automate extraction of daily and annual mean values & export to CSV (Fig 4)
• Produce final table of statistics to be published by Bank of England (Fig 5)

After completion, I designed and developed a climate change data portal, specifically to distribute the data aggregated and processed for this task. Link to portal