Qualifier Seminar Mr. Muhammad Imran

Scale and Uncertainty in Spatio-Agricultural Models within Spatial Data Infrastructure

Integrated Assessments address the problem that cannot be solved by a single discipline of science. The researchers from different disciplines need to work collaboratively and synthesize unique knowledge to design and implement effective solutions. For instance, a sustainable agriculture system cannot be accurately modeled without integrated assessments across disciplines such as biophysics, economy and environment. An Integrated Assessment Scenario (IAS) is a tool that helps decision-makers to improve their understanding of, for instance, interrelated complex agricultural processes at various spatial scales, in different thematic domains. In environmental sciences, the computational models have been designed to qualify or quantify the observed entities and processes in space and time at a certain scale. Technically, an IAS is realized as a workflow involving multiple computational models, data inputs and outputs. In geo-information science, spatial and temporal scale of data and modelling scale are two variants of IAS which are prominent in the context of problems associated with interoperability and data quality, such as how to dynamically integrate diverse IAS components at different domains and scales and how to quantify and communicate the uncertainty associated with dynamic interactions in the processes in such integration.

This research aims at providing more intelligent, flexible and interoperable IAS frameworks for knowledge integration at scales, by providing the basis for reasoning with quality and uncertainty. It will address the question of how different IAS components such as diverse data sets, simulation models and solutions from different disciplines and different working groups can effectively integrate their knowledge for solving a given problem of an integrated assessment and what methods could be applied for facilitating this process. The scientific philosophy of this research is to provide:

• Reasoning with metadata on dynamic integration of the diverse datasets and simulation model components in IAS.
• Reasoning with uncertainty on accurate linking of datasets and simulation model components in IAS.

This scientific advancement will be highly useful in conducting more precise integrated
assessments in a SDI framework, particulary in the Sub-Saharan African (SSA) agriculture, which pertains to:

• Spatio-agricultural model scaling at different spatio-temporal scales and disciplines.
 
In the technological perspective, an Open Architecture has been proposed for IAS, that will allow to publish the process models as web services with semantic catalog entries along with developing and storing the model descriptions as ontologies in the database. A reasoning mechanism which may be an agent can itself discover such published models according to the descriptions of an intended scenario. Ideally, a reasoning mechanism should be able to generate code and quantify uncertainty from model descriptions in an automated way for the model run, following the declarative modeling paradigm. Such ontology-based solutions can be the basis of developing highly flexible frameworks for IAS, in which the reasoning with quality and uncertainty can be in place to integrate the data sources (data services), data transformation mechanisms (processing services), and models (simulation services) at different scales and disciplines, according to the descriptions of models and end-user scenarios.

previous page
more events