Earth Observation Data Processing Cloud Requirements and Research

Earth Observation Data Processing Cloud: Requirements and Research Issues Nitant Dube Space Applications Centre, ISRO India

Sub-meter Agile View R E S O L U T I O N Background Technology Roadmap Cloud Computing Grid Computing 1 -3 m Hyper Spectral Distributed and Cluster Based Computing Component Based Architecture 10 m Stereo Monolithic Computing 30 m Multi. Spectral Service Oriented Architecture Limited Digital Processing B&W Photographs Modular Programming Visual Interpretation 1 st Generation 1960 -1972 2 nd Generation 1972 -1986 3 rd Generation 1986 -1997 Earth Observation Generations 4 th Generation 1997 -2010 5 th Generation 2010 -2020

Satellite Data Processing Overview Data Acquisition System Information Management System User Request Browse & Accession Catalogue Product Archive Data Products System Products Value Added Data Products System Products Data Quality Check Data Quality Evaluation

Requirements of Earth Observation Data Processing (Infrastructure) Capability to migrate legacy and operational code to new infrastructure, without requirements for change in software Reprocessing of old data, requires huge computational loads EO data processing usually generates burst loads (LEO, GEO), capability to optimally utilize the resources without addition of extra complexity in the software Improved Reliability in operations Time Series Processing and Analysis of Earth Observation Data

Cloud Computing Provides computing and resources (both hardware and software) on demand without worrying about the complexity and details of the underlying infrastructure Allows Systems to scale-up and scale-down (both capacity and functionalities) based on requirements Strengths Weakness Allows users to create virtual organization, where resources can be optimally used. Performance, reliability, SLA, Data control, Standardization, Inter operability

Requirements of Satellite Data Processing (Software) Capability to Search Service not only based on Name or provider but also on parameters to meet the functional and quality requirements Needs a framework where new Research can be used by existing applications and existing services can be used by new software Support for Versioning in services, whenever services are enhanced and as new versions are released, there should be capability to use either the latest version of the service or fixed version based on user requirement Need for a Service Evaluator, which can verify the quality parameters based on data set submitted by the user. Ease of Use for service composition and option for the user to change the composed application

Service Oriented Architecture 1 Service Broker Describe • Loose Coupling 2 Discovery • Capability to • Discover • Describe Client Service Provider 3 Uses • Deliver • Compose

Layered Architecture User EO-Cloud Radiometric correction, Geometric Correction, Atmospheric correction, Orthrectification… Software as a Service Platform as a Service Infrastructure as a Service storage CPU Erdas, Arc. Info, OSSIM, Orfeo Tool. Box, … Storage, Computational nodes, network

EO-cloud Application Scenarios

Data/Information products using Multisource EO data Virtual Data/Information Processing System EO-Cloud Satellite Data Model Forecast Reference Images • EO-Cloud can provides access to multiple data sources for quality improvement during product generation.

Advanced data visualization Virtual Reality Applications EO-Cloud DEM Images Geo-Spatial Layers • Will Provide Access to Dynamic, Distributed and Multi -Source Geo-Spatial Data for Virtual Reality Applications

Algorithm Development and Fine-Tuning of Products Using Multi-Satellite Data Source Algorithms Algorith Development Framework EO-Cloud Open. Street Map Optical Data Microwave Data • EO-cloud can provides access to multiple data sources, Computational Resources and existing Software for New algorithm development.

Research Issues Cloud Workflows for Satellite Data Processing Economy of Scale Cloud Interoperability

Thanks nitant@sac. isro. gov. in