TY - GEN
T1 - A spatial time series forecasting for mapping the risk of COVID-19 pandemic over Bandung Metropolitan Area, West Java, Indonesia
AU - Manessa, Masita Dwi Mandini
AU - Kamil, Ridwan
AU - Setiaji, Setiaji
AU - Ningrum, Ida
AU - Suseno, Weling
AU - Rahmayanti, Ira
AU - Zulkarnain, Faris
AU - Ardiansyah, Ardiansyah
AU - Lesmini, Indah
AU - Tasdiq, Rahmat Hidayatulloh
AU - Moe, Idham Riyando
N1 - Publisher Copyright:
© 2020 SPIE
PY - 2020
Y1 - 2020
N2 - West Java is in the five line on the list of provinces in Indonesia with the most COVID-19 cases, as Bandung Metropolitan Area (BMA) is the second most densely populated showing the highest number after Jakarta Greater Area. Bandung Metropolitan Area consist of Bandung City, Cimahi City, Bandung Regency, and West Bandung Regency. Then, an intense movement of people created between the connected city and regency. Bandung City became the epicenter of movement BMA, since it is the province capital city, business, and education center. This fact, putting BMA at the highest risk not only for the pandemic but also socioeconomic issues. The spatial time series risk forecasting information is an essential for the decision-maker to develop a day by day policy aimed for combating the COVID-19 pandemic issue. In this study, the pandemic risk is calculated by combining vulnerability, hazard, and geodemography information. Infimap provides the People in Pixels geodemographic data, added not only the exposure of population distribution to COVID-19 but also the ratio of age. Beside those data, the daily distribution of COVID-19 cases, network data, business point, health facility point, residentials area, geodemographic (People in Pixels), and daily COVID-19 Community Mobility Reports is also been used in this study. The daily vulnerability and hazard data created since the first case on March 4th until August 21st. The hazard area is create based on the expected travel area of positive COVID-19 patient. While the vulnerability area is create using Spatial Multi Criteria Analysis (SMCA) of following data: service area of hospital, groceries (local market), and workspace. Further, the time series data of hazard and vulnerability area was inputted to develop the forecasting model based on the machine learning pipeline of Gaussian algorithm. As a result, this study shows the possibility to predict the future risk area of COVID-19 until the next 100 days condition, based on spatial timeseries forecasting model.
AB - West Java is in the five line on the list of provinces in Indonesia with the most COVID-19 cases, as Bandung Metropolitan Area (BMA) is the second most densely populated showing the highest number after Jakarta Greater Area. Bandung Metropolitan Area consist of Bandung City, Cimahi City, Bandung Regency, and West Bandung Regency. Then, an intense movement of people created between the connected city and regency. Bandung City became the epicenter of movement BMA, since it is the province capital city, business, and education center. This fact, putting BMA at the highest risk not only for the pandemic but also socioeconomic issues. The spatial time series risk forecasting information is an essential for the decision-maker to develop a day by day policy aimed for combating the COVID-19 pandemic issue. In this study, the pandemic risk is calculated by combining vulnerability, hazard, and geodemography information. Infimap provides the People in Pixels geodemographic data, added not only the exposure of population distribution to COVID-19 but also the ratio of age. Beside those data, the daily distribution of COVID-19 cases, network data, business point, health facility point, residentials area, geodemographic (People in Pixels), and daily COVID-19 Community Mobility Reports is also been used in this study. The daily vulnerability and hazard data created since the first case on March 4th until August 21st. The hazard area is create based on the expected travel area of positive COVID-19 patient. While the vulnerability area is create using Spatial Multi Criteria Analysis (SMCA) of following data: service area of hospital, groceries (local market), and workspace. Further, the time series data of hazard and vulnerability area was inputted to develop the forecasting model based on the machine learning pipeline of Gaussian algorithm. As a result, this study shows the possibility to predict the future risk area of COVID-19 until the next 100 days condition, based on spatial timeseries forecasting model.
KW - COVID-19
KW - Geodemographic
KW - GIS
KW - Hazard
KW - Risk
KW - Spatial time series
KW - Vulnerability
UR - http://www.scopus.com/inward/record.url?scp=85093675297&partnerID=8YFLogxK
U2 - 10.1117/12.2572536
DO - 10.1117/12.2572536
M3 - Conference contribution
AN - SCOPUS:85093675297
T3 - Proceedings of SPIE - The International Society for Optical Engineering
BT - Earth Resources and Environmental Remote Sensing/GIS Applications XI
A2 - Schulz, Karsten
A2 - Michel, Ulrich
A2 - Nikolakopoulos, Konstantinos G.
PB - SPIE
T2 - Earth Resources and Environmental Remote Sensing/GIS Applications XI 2020
Y2 - 21 September 2020 through 25 September 2020
ER -