Natural disasters can have long-lasting financial consequences for vulnerable populations such as farmers, but disparate disaster impacts make the accurate and timely deployment of limited resources such as disaster funding and relief very difficult. This project will develop an open-source suite of cyberinfrastructure tools, called the Decision Engine for Socioeconomic Disaster Risk (DESDR), to fill data voids by collecting and cleaning disaster risk data directly from affected populations. As well, the project will combine those data with satellite data to provide farmers and others with data and tools to make better informed decisions about disaster risk management.
Existing approaches for disaster detection, mapping, and prediction primarily rely on satellite and sensor data that are susceptible to errors and do not measure how disasters directly affect regions and individuals. This project will develop open-source software infrastructure employing data from rural populations, and then use the data to build more accurate disaster risk models. The cyberinfrastructure, DESDR, will provide a scalable, customizable data collection platform using mobile messaging services that take local incentives and community norms into account, and will be used to gather data on from affected communities. A data visualization and cleaning platform will enable local partners and researchers to cross-reference the data with satellite and sensor data sources, and to identify and clean data errors. A database architecture will store the data in an interoperable format. And, a web-based interface will provide government agencies, policy makers, researchers, and other stakeholders the ability to interactively create and back-test disaster risk models.
DESDR will be disseminated as open-source, ready-to-deploy software to a user community of governmental meteorological agencies, humanitarian program officers, insurers and affiliated agricultural and social scientists. By combining these tools into an integrated, extensible process cyberinfrastructure, DESDR will directly involve vulnerable populations in the design of solutions, enabling disaster risk managers to scale up critical relief programs to reach multitudes of farmers and others while providing an unprecedented voice to the project beneficiaries.
iKon has been deployed twice, and collected thousands of responses from farmers:
The Reptile Data Cleaning System has been used to clean hunderds of survey records by local users in two field tests:
Our DBT-based Insurance Design Tool has been field tested in Zambia and Ethiopia:
WhatsApp farmer crowdsourcing tool Survey your Way . Developed as part of the NSF DESDR grant at Columbia University: