MOOD Maps of Google community mobility change during the COVID-19 outbreak

The MOOD project (MOnitoring Outbreak events for Disease surveillance in a data science context. H2020) has geo-referenced the data Google has published as a series of PDF files presenting reports on national and subnational human mobility levels relative to a baseline data of late January 2020. The details and the PDF files can be found at https://www.google.com/covid19/mobility/.<br>More detail on these files can be found at https://www.moodspatialdata.com/humanmobilityforcovid19 <br>The first set of data were released on April 2 2020 and have been revised weekly since then. The maps now utilise the CSV data released by Google. Please note that the maps figures use a mean of the previous three days, while the Google PDFs use a single days data so there will be differences between values in our maps when compare to the Google PDFs.<br>The authors have extracted the majority of these data into a series of excel spreadsheets. Each worksheet provides the data for % change in numbers of records at various types of location categories illustrated by: retail and recreation, grocery and pharmacy, parks and beaches, transit stations, workplaces and residential (columns f to K). A second set of columns calculates the difference of each value from the mean values for each category (columns L to P) Columns A to E contain geographical details. Column Q contains the names used to link to a mapping file.There are separate worksheets for the date of the data from each dated release (e.g. 2903, 0504 etc.) and separate worksheets calculating the changes between specific dates.<br>A second spreadsheet has been added calculating the 3 day moving mean of each day from the 15th of February. Each day is referenced by the Gregorian calendar day count. So day 48 = Feb 17th.<br>The maps (for EU &amp; Global) display these data. We provide 600 dpi jpegs of the Global (“WD”) and European (“EU”) mapped values at the latest date available, for each of the mobility categories: retail and recreation (“retrec”) , grocery and pharmacy (“grocphar”) , parks (“parks”) , transit stations (“transit”), residential (“resid”) and workplaces (“work”). We also provide maps of the changes from the previous week (“ch”).<br>All data extracting and subsequent processing have been carried out by ERGO (Environmental Research Group Oxford) on behalf of the MOOD H2020 project. Data will be periodically updated. Additional maps can be obtained on request to the authors.

MOOD项目（数据科学框架下的疾病暴发事件监测项目，属欧盟H2020计划资助项目）已对谷歌发布的一系列PDF报告数据进行地理配准，这些报告以2020年1月末的基准数据为参照，展示了国家级及次国家级的人类流动水平。相关详情及PDF文件可访问https://www.google.com/covid19/mobility/获取。
关于这些文件的更多细节可访问https://www.moodspatialdata.com/humanmobilityforcovid19查阅。
首批数据集于2020年4月2日发布，此后每周进行修订。当前的可视化地图采用谷歌发布的CSV格式数据，需注意：本数据集地图使用过去三日的平均值进行计算，而谷歌原始PDF报告仅采用单日数据，因此本数据集地图与谷歌PDF报告中的数值存在差异。
项目团队已将大部分数据提取至一系列Excel电子表格中。每个工作表均包含不同场所类别的记录数量变化百分比数据，涵盖零售与休闲、食品杂货店与药房、公园与海滩、公共交通站点、工作场所及居民区共六大类别（对应表格第F至K列）。第二组列用于计算各类别数值与对应类别平均值的差值（对应表格第L至P列）。第A至E列为地理详情信息，第Q列为用于关联映射文件的名称字段。每个数据集发布日期（例如2903、0504等，即月日格式）对应单独的工作表，同时另有单独工作表用于计算特定日期之间的变化量。
新增的第二份电子表格用于计算2月15日起每日的3日移动平均值，其中每日以公历累计天数进行标识，例如第48天对应2月17日。
欧盟（EU）与全球（WD）范围的可视化地图展示了上述数据。本数据集提供最新日期下，全球（WD）及欧盟（EU）范围内各流动类别的600 DPI JPEG格式地图，涵盖的流动类别包括：零售与休闲（"retrec"）、食品杂货店与药房（"grocphar"）、公园（"parks"）、公共交通站点（"transit"）、居民区（"resid"）及工作场所（"work"）。同时还提供相较于前一周的变化量地图（"ch"）。
所有数据提取及后续处理工作均由牛津大学环境研究组（ERGO）代表MOOD H2020项目完成。本数据集将定期更新，如需额外地图可联系作者获取。