How have the lockdown measures impacted mobility in the Balearics?

The IFISC (CSIC-UIB) is heading a CSIC project using anonymised mobile data to study the effectiveness of the lockdown on the spread of COVID-19

The Institute for Cross-Disciplinary Physics and Complex Systems (IFISC, UIB-CSIC) is analysing the changes to mobility in the Balearic Islands arising from the exceptional lockdown measures to fight the spread of COVID-19. The study is part of a Spanish National Research Council (CSIC) project coordinated by the IFISC (CSIC-UIB) and the Blanes Centre for Advanced Studies (CEAB-CSIC). The project is using data science techniques to verify the effectiveness of the lockdown measures taken to halt the spread of COVID-19.

The Balearic mobility reports use aggregate data (i.e. taking into account an entire working day) provided by Kido Dynamics. The data are based on mobile records from Orange, representing around 25% of the Spanish market, and analyse mobility across two categories: within each municipality and between municipalities.

The reports are released weekly and compare data from different days: Wednesday 4th March (before the state of emergency), Wednesday 18th March (after the introduction of the state of emergency and the first round of restrictions limiting movement with justifiable cause) and Wednesday 1st and 8th April (after the introduction of the second round of measures that closed down all non-essential activities). In order to analyse how mobility has changed, the relative difference between the selected days is calculated. The first report issued by the IFISC (CSIC-UIB) offers a similar analysis for 4th, 6th, 18th and 20th March, all dates before the second round of measures were introduced (29th March).

In terms of mobility between municipalities, there is a major contraction (50% on average) for journeys on 18th March when compared to 4th March. This reduction intensifies on 1st April, with the second round of measures (around 70-80%). The results show a major decline in trips between municipalities in the Balearic Islands. Given the data source, journeys cover all types of transport methods, including on foot, and the sections between stops are counted separately.

Studying mobility within each municipality is more complex. On the one hand, there was a 20% average drop in mobility after the first round of measures, increasing to 40-50% after the second round of restrictions were put into place. On the other, some inland municipalities in Majorca have seen higher internal trip numbers and this increase, already noted after the introduction of the first round of measures, in some cases rises to 50%. This is due to the fact that before the measures, inhabitants would travel outside the area to work and now remain in the municipality most of the day, meaning any necessary trips for shopping and agricultural activities are clearly within the municipality, leading to high relative increases, albeit not in net values. A similar effect has been observed in rural areas on the mainland.

Finally, journeys between municipalities on the different islands are analysed to see whether the changes follow a similar pattern. Specifically, trips from and to Palma have declined by 60% (18th March) and 70-75% (1st and 8th April). The decline has been between 50-75% in Ibiza and Minorca.

All the reports with details for each municipality, as well as others for each province around Spain, are available on the Data Analytics @ IFISC website. This unit has been set up by the IFISC to provide data mining, and big data analysis and modelling services to the CSIC, UIB and external institutions.

Mobile data to study lockdown effectiveness

These reports are part of a wider CSIC research project led by the IFISC (CSIC-UIB) and the CEAB (CSIC). The results will be essential to improve social distancing strategies to be adopted in future outbreaks of this or any other disease. A multidisciplinary team of computing, demographics, physics and transport study experts has been put together to undertake the research and analyse big, high resolution data obtained from map servers and mobile operators. The data explain how mobility and social contact have changed since the start of the lockdown.

The research team simulates different scenarios or strategies for social distancing and decision-making through the obtained data. The results are key to deciding on whether a stricter lockdown is initiated or on planning a safe, effective end to lockdown.

In order to achieve its objective, the project includes several parallel stages. First, mobility is characterised thanks to the contribution from different data platforms, e.g. social networks and mobility patterns reflected in aggregate mobile records. A second aspect is the change in people's behaviour due to their perception of risk. Mobile surveys and applications are run as part of the project to quantify the changes; in addition, an attempt is made to estimate the level of adherence amongst the population to the personal protection measures, as well as the changes in quality and quantity of social contact. This information is key to understanding the contagion process.

The project will use artificial intelligence and data science tools, and include real-time big data on human mobility, geolocation surveys and computing models. It will represent a new way to look at epidemiology, combining computational epidemiology, digital demographics and human mobility models. All of this will enable observation of how containment measures have altered people's mobility and behaviour.

A second long-term aim is to establish the beginnings of a computational epidemiology network in Spain (something that already exists in other countries) and a set of interoperable analytical tools based on epidemiology theory, data science and artificial intelligence to feed decision-taking in future epidemiological crises that could re-occur in our globalised, interconnected world.

The information and models to be developed in this research will be made publicly available for future use in line with an open-access data model based on FAIR principles (Findable, Accessible, Interoperable, Reusable).

In addition to the IFISC (CSIC-UIB) and the CEAB, teams from the Institute for Economics, Geography and Demographics (IEGD-CSIC), the Cantabria Institute of Physics (IFCA-CSIC) and the National Biotechnology Centre (CNB-CSIC) are taking part in the project, as well as scientists from the Pompeu Fabra University and the National Epidemiology Centre-Institute of Health Carlos III (ISCIII).

Source: IFISC (CSIC-UIB)

Related news

Publication date: Mon Apr 20 09:29:00 CEST 2020