Mapping seasonal demography and mobility for malaria elimination

Research Project

Project/Area Number	20K10447
Research Category	Grant-in-Aid for Scientific Research (C)
Allocation Type	Multi-year Fund
Section	一般
Review Section	Basic Section 58020:Hygiene and public health-related: including laboratory approach
Research Institution	The University of Tokyo
Principal Investigator	新井亜弓東京大学, 地球観測データ統融合連携研究機構, 特任研究員 (10788574)
Co-Investigator(Kenkyū-buntansha)	金杉洋東京大学, 空間情報科学研究センター, 客員研究員 (00526907) ウィタヤンクーンアピチョン東京大学, 空間情報科学研究センター, 特任助教 (90726407)
Project Period (FY)	2020-04-01 – 2024-03-31
Project Status	Granted (Fiscal Year 2020)
Budget Amount *help	¥4,160,000 (Direct Cost: ¥3,200,000、Indirect Cost: ¥960,000) Fiscal Year 2023: ¥130,000 (Direct Cost: ¥100,000、Indirect Cost: ¥30,000) Fiscal Year 2022: ¥1,820,000 (Direct Cost: ¥1,400,000、Indirect Cost: ¥420,000) Fiscal Year 2021: ¥1,040,000 (Direct Cost: ¥800,000、Indirect Cost: ¥240,000) Fiscal Year 2020: ¥1,170,000 (Direct Cost: ¥900,000、Indirect Cost: ¥270,000)
Keywords	mobile phone data / human mobility / developing country / malaria / public health
Outline of Research at the Start	This study will first address how sociodemographic factors vary with movement, and map demographic variables important to malaria transmission and risk throughout Mozambique during each month of the year. This study will combine seasonal demography and movement predictions with monthly malaria risk maps to predict populations at-risk of malaria exposure in high transmission months.
Outline of Annual Research Achievements	Work Package I: Linking seasonal demography and mobile phone patterns to map seasonal demography Data access is already secured. We already set up data processing environment in the premise of the ICT regulator who facilitated mobile phone data collection from three mobile network operators (MNOs). The mobile phone data of the three MNOs cover more than 90% of the market so we consider that the data well represent the general population of study site. We examined the correspondence of residential population observed from mobile phone data with known population data (census). The result shows that the population distribution estimated from mobile phone data in terms of residential population is relevant for this study (we used the data of July 2019). We conducted the analysis on mobility patterns by different seasons. We estimated the population distribution at different times of days to examine how population distributions change seasonally; we also examined changes in residential locations. It enables us to indicate the temporal changes in residential location. It can be used as a proxy for seasonal migration. To compare the seasonal characteristics of mobility patters, we employed the following mobility metrics: OD matrices, average distance traveled, and radius of gyration. Estimated demographic attributes using our existing model where we found that the model is not robust under COVID-19 setting as people's mobility patterns are different from usual patterns. Work Package 2: Focusing on high transmission seasons and extending to other countries Data collection is ongoing.
Current Status of Research Progress	Current Status of Research Progress 3: Progress in research has been slightly delayed. Reason We found that our existing model for estimating demographic attributes is too sensitive to the changes in time spent at home and work places, which are key parameters in the model. The estimation result is not encouraging. We are considering to taking a different approach for estimating the attributes of mobile phone users.
Strategy for Future Research Activity	1) Attempting to change data period for this study due to the impact on mobility patterns under COVID-19 We already have access to mobile phone data of July 2019 (dry season) and February 2020 (rainy season) where the mobility patterns of February 2020 is affected by social-distancing policy. We are trying to obtain February 2019 as the alternative to February 2020 where we expect to be able to extract usual seasonal patterns without any effects of COVID-19. 2) Possibility of developing a different approach to infer the sociodemographic attributes of mobile phone users We are considering that we may use a different way of classifying population groups by sociodemographic attributes. In our original plan, our model estimate demographic attributes based on routines in mobility patterns. However, parameters used for the estimation are closely related to mobility patterns and may not work effectively under COVID-19. Instead, we plan to use satellite images to estimate socioeconomic groups by the pattern of buildings and human settlements.