New Location Data Available for COVID-19 Recovery Analysis

More than two months have passed since the World Health Organization declared the outbreak of COVID-19 a worldwide pandemic. Governments from all over the world have had to take unprecedented measures to force people to stay at home in order to prevent the coronavirus from spreading further. While stay-at-home measures have proved to have been effective in controlling the reproduction ratio of the virus and have helped to save hundred of thousands, if not millions, of lives; significant collateral damage has been caused in the economy at global, national and local levels.

After weeks of strict lockdown, some governments are now starting to implement plans of progressive re-opening of economic activities while slowly lifting stay-at-home orders. In the same way the coronavirus outbreak caused more havoc in some regions than others, these plans to transition into a new normal include phases that will not be implemented simultaneously in all parts of a country.

Location Intelligence has already played a critical role in understanding the nature of the virus spread and empowered administrations with tools to better respond in such uncertain times. As we move into the next phase and during the upcoming weeks, it will earn a new degree of prominence by helping businesses and administrations to make more informed decisions when adapting to the new status quo and optimising resources for the fastest possible recovery.

New public datasets available in the Data Observatory

Continuing our effort to provide smooth access to ready-to-use and up-to-date location data, so data scientists and analysts can focus on generating insights to help their institutions navigate this fast changing environment, the team at CARTO has kept working hand in hand with our network of data providers to include new public datasets in the Data Observatory.

Infection rate and risk factors

One of the new datasets that we are now making available in the Data Observatory has been curated by Doorda, a provider of demographic, health and housing data for the UK market. This dataset provides estimates for a set of risk factors (age, number of household residents, smoking habits, etc.) and COVID-19 infection rates at a range of local geographies. Thanks to its weekly updates, this dataset allows for a detailed socio-demographic and socio-economic analysis of the impact over space and time of the coronavirus in the UK.

Social media sentiment

The COVID-19 outbreak has dramatically changed our daily activities, forcing us to modify our routines and the way we carry out (the many times necessary) social interactions. All these changes can be identified online if one can analyse how people have communicated across social networks and digital platforms during the last couple of months. Spatial.ai, a US-based provider of geosocial segments and sentiment data, has built a dataset analysing the sentiment of social media posts mentioning COVID-19. Sentiment indicators, calculated using the VADER (Valence Aware Dictionary and Sentiment Reasoner) model, are then aggregated by US county and updated weekly. Leveraging social media sentiment is a novel and insightful method to analyse the impact the virus is having on people’s moods and behaviors and how these evolve in conjunction with the epidemiological indicators and the application of the different degrees of shetler-at-home orders at county level.

Human mobility data

Human mobility has been one of the most affected factors during the COVID-19 crisis, since limiting it has proved to be one of the most efficient ways to stop the virus from spreading further. As seen by the abundance of press articles on the topic, the analysis of human mobility data, based on the aggregation of location signals captured by certain apps installed on mobile phones, has been one of the most powerful resources used by journalists, researchers, and administrations to measure the efficacy of social distancing initiatives. This data is now also being used to help analysts understand the retail comeback by measuring changes in visitation patterns at retail areas.

Apple’s Mobility Trends Reports provide daily indicators showing the relative volume of direction requests in Apple Maps per country, region or city compared to a baseline volume on 13th January 2020. In order to allow users to enrich Apple’s reports with other sources of location data, CARTO’s team has undertaken the effort of geocoding the different geographic regions detailed in the data.

CARTO’s human mobility data partners Unacast and SafeGraph have been empowering the work of research institutions, nonprofit organizations and public agencies by giving them free access to special COVID-19 related datasets that measure mobility changes at different spatial aggregations. Thanks to our close collaboration with them, these data products are now also available via CARTO’s Data Observatory for those institutions participating in their respective “Data For Good” initiatives as well as under commercial licenses for any for-profit organization. Note that CARTO’s platform is also available to public and private sector organizations doing research on the coronavirus pandemic via our grants program.

Unacast’s Social Distancing indicators are currently available at regional level for a set of countries, including the US, UK, Brazil, Mexico, and France. As a reference, the US metrics are available aggregated at the county level and provide daily measurements on the change in average distance travelled and the change in probability of human encounters.

SafeGraph has made available its Social Distancing Metrics dataset with metrics aggregated at the block group level, updated daily, and focused on the US. SafeGraph data provides information on the median distance traveled from home, the ratio of devices spending all day at home, and the distribution on the time others are spending out of their homes. Additionally, as a resource for researchers on COVID-19, SafeGraph is also providing weekly updates of a “light” version of their Patterns dataset.

This notebook we are sharing via Google Colab illustrates how to access some of the aforementioned datasets that are now publicly available from CARTO’s Data Observatory, as well as from our public project in Google BigQuery.

In need of support to apply Spatial Analysis and Spatial Data Science in the fight against COVID-19?

Click here to find out more

EU Flag
This project has received funding from the European Union’s Horizon 2020 research and innovation programme under grant agreement No 960401.
About the author
Javier Pérez Trufero

Javier is CARTO's Head of Data, running the company strategy with respect to third party data offerings and data science activities. Javier's responsibilities at CARTO range from establishing and coordinating new alliances with top-class data providers such as Vodafone and Mastercard, to contributing to the definition of CARTO's data products and data science projects.

More posts from Javier Pérez Trufero
About the author
Alejandro Polvillo

Data. Machine Learning. Deep Learning. Software Engineer. Data Scientist at CARTO.

More posts from Alejandro Polvillo
About the author
Álvaro Arredondo

Data Scientist at CARTO

More posts from Álvaro Arredondo

Related Posts

Ready to optimize your territories with Location Intelligence?

Close circle icon

Contact us

Please fill out the below form and we'll be in touch real soon.