Disparate Impacts Require Deliberate Solutions

When it rains, it pours unevenly — both literally and figuratively. The effects of change do not fall proportionately across genders. For example, we know that women and girls are much more likely to bear the costs of climate change. 

Climate change poses a profound threat to women and girls, who bear a disproportionate burden of its consequences. UN reports estimate that 80% of those displaced by climate change (in the 59 low- and middle-income countries analyzed) are women and girls, which escalates their risks. Furthermore, the mortality rate for women and children during natural disasters can be up to 14 times higher than that of men. This heightened vulnerability is often due to socio-economic factors: women's high dependence on climate-sensitive livelihoods, such as small-scale agriculture, makes them more susceptible to losing income due to floods or droughts. 

While extreme weather exacerbates inequities, women face disparate impacts with any type of change or uncertainty. Social and economic change also impacts women more severely. For example, a World Bank study found that a 1% increase in the male unemployment rate is associated with an increase in the incidence of physical violence against women by 0.5 percentage points, or 2.75%. This is consistent with the financial and psychological stress generated by unemployment. 

Understanding these challenges is the first step to solving them, or at least mitigating their impact. Fortunately, we have the tools to help do that, and historical examples to guide us. First of all, we need data, and in the age of AI, we need a lot of it. We need it to be accurate, timely and representative of the populations we’re trying to understand — these are the goals of Snowflake’s End Data Disparity movement and the new Data for Good data label in Snowflake Marketplace. But we also need a means of analyzing the data. Maps provide a valuable tool. 

If a picture paints a thousand words, maps are off the charts

There is a long history of using maps to illustrate patterns in data and find answers to challenging questions. One of the first examples of early thematic mapping comes from London physician John Snow in 1854. Starting with a map of a London neighborhood including streets and water pump locations, Snow mapped out the incidence of cholera deaths. Overlaying this data, a pattern emerged, centered around one particular pump on Broad Street. Upon removal of the pump’s handle, new cholera cases stopped. It turns out the Broad Street pump was near a cesspit under the home of the outbreak's first cholera victim.

Colors enhance illustration, helping to distinguish between different types or values in the data. For example, colors illustrate the effect of the potato famine in Ireland from 1841 to 1851 in the map to the left. Dark brown represents areas with an over 30% drop in population. But the three areas in green also jump out. These are the largest cities in Ireland: Dublin, Cork and Belfast. These cities were likely less dependent on potatoes for food and as a source of income. Why? Diverse economies? Diverse diets? Proximity to ports and trade? Maps reveal patterns and improve understanding.

These are just a few historical examples of how maps paint even more than a thousand words. They offer insights into challenges and suggest solutions. With more data to overlay on them today, maps are an even more powerful tool.

Charting a new course: Snowflake and the Ordnance Survey

Snowflake has teamed up with the Ordnance Survey to bring geospatial tools and data together. The Ordnance Survey is Great Britain’s national mapping service, with a heritage stretching back over 230 years. It has evolved from traditional mapmakers into a data-driven organization that provides critical geospatial infrastructure and applications. 

OS manages the National Geographic Database — the authoritative source of location data in Great Britain. It builds digital products and APIs for developers and businesses. On the international stage, it partners with governments globally. And it also empowers outdoor enthusiasts through consumer products such as the OS Maps App.

At the recent Big Data London conference, Snowflake and OS presented jointly on the power of using mapping to visualize data, highlighting two examples. 

Mitigating the impact of flooding: Rebekah Spratt, Geospatial Graduate at OS, presented her research on building climate resilience in Kinshasa through a project with the UK Government Centre of Excellence, Green Cities, Infrastructure & Energy. The project addressed flood risks in one of Africa's largest cities, home to nearly 18 million people and experiencing rapid growth. Sitting on the banks of the Congo River, the city has experienced increased flooding due to climate change. 

Combining data on environmental risk, urban growth and vulnerable populations, the project used disaster and risk modeling and spatiotemporal analysis to deliver an accessible tool for disaster management designed for city planners and policymakers.

Preventing violence against women: Jonathan Allsup, Geospatial Specialist at OS, then presented a recent collaboration between OS and Snowflake that had developed a new data-powered example approach for the UK government's program Violence Against Women and Girls (VAWG).

The new Crime Reporting and Intelligence Platform uses OS data from the National Geographic Database — buildings, land use, urban infrastructure such as streetlights and data from OS Open Greenspaces — to analyze incidents of violence against women and girls in urban spaces. The demonstration tool, built using Streamlit, reports each incident’s proximity to different elements of the urban environment, visualized on a map, and presents both spatial and temporal patterns. For example, the results indicate that incidents of violence are more likely to occur next to retail outlets than residential locations and are correlated with a lack of lighting. The resulting spatiotemporal insights about crime in the West Midlands of the UK help emergency services better allocate resources to areas of concern and ultimately prevent future incidents.

These examples illustrate how location data underpins nearly every policy decision — from where to install additional lighting or dispatch emergency services to how to respond to floods or other extreme weather. Why? Because everything happens somewhere. Mapping out data reveals patterns and offers insights into potential solutions. 

Now, Ordnance Survey data — with 11 data sets, including Boundaries, Open Roads, Open Rivers, Open Names, Open Greenspaces and more — is available on Snowflake Marketplace. 

Whether assessing climate risk in financial services or identifying vulnerability hotspots for police forces, location data helps us understand where things happen — and more importantly, can offer insights into why. Where the impact of events is disparate across populations, these insights enable deliberate policies and solutions.