WP3 Multi-resolution dynamic sampling and integration framework

Lead: GFZ

Reliable and cost-effective monitoring of vulnerability and risk, as well as of recovery and reconstruction processes can be achieved by the careful integration of different data-sources. Prioritising data-collection efforts with respect to the estimated level and extent of hazard and exposure, and according to end-users and stake-holders constraints and requirements, is the baseline upon which a dynamic, multi-resolution  sampling framework can be developed. Sampling refers to both supervised and unsupervised information extraction from remote-sensing data, and to in-situ data collection or direct observations. Monitoring of the spatio-temporal variability of indicators for vulnerability, recovery and reconstruction is dependent not only on the amount and quality of the information upon which the assessment is made, but also on the tools and methodologies used to analyse, manage and present this information. Spatio-temporal changes need to be properly integrated into a sound, comprehensive conceptual and methodological framework, which is able to deal with multi-dimensional data coming from different sources, at varying scales and changing over time.

The main goal of this work package is to provide a comprehensive multi-resolution dynamic sampling and integration framework, which consists of both methodological guidelines and software solutions to efficiently model geographically-distributed, multi-dimensional data. Central to the work package is the concept of multi-resolution, dynamic sampling and integration, as a way to adaptively and optimally collect  information, which is typically resource- or time-expensive. Furthermore, this work package aims at providing a knowledge life-cycle management system to optimally  merge(which accommodates uncertainly) newly acquired information, taking into  account pre-existing knowledge, and disposing vintage or unreliable information.

The key objectives of this WP are:

  • Collection and storage of data from different sources, at varying scales and changing over time into a geospatial-database largely based on existing free, open-source solutions.
  • Developing strategies to derive focus maps and setting up a dynamic sampling framework that allows for  effective and optimal information collection with variable resolution. Focus maps focus the sampling framework on the most critical areas to be monitored, tailoring the knowledge collection efforts in terms of an end-user's priorities and available resources.
  • Setting up of a knowledge-based information life-cycle management solution.