In a crisis, up-to-date information is one of the most important commodities for decision-makers. Remote sensing data have been instrumental in regional scale damage detection and recovery progress monitoring after significant disasters. However, using remotely sensed data to support an emergency response requires not only the availability of hardware, software, and manpower to process and analyze the data, but also the time to stage the datasets that are required for analyses. Additionally, the volume of remote sensing data that needs to be processed to detect temporal changes accurately in a terrestrial or oceanic ecosystem can easily exceed several terabytes, even for a small region. This is because emergency response requires the use of well-calibrated remotely sensed data products, like those that are generated by the MODIS (Moderate Resolution Imaging Spectroradiometer) Adaptive Processing System (MODAPS). These data sets are stored and distributed by the Level 1 and Atmosphere Archive and Distribution System (LAADS), both located at Goddard Space Flight Center (GSFC), and are necessary to create the custom data products that are needed and used for emergency management situations. Generally, the MODIS datasets are downloaded from GSFC, stored at the user’s facility, and then processed locally. This approach is standardly used by researchers worldwide.

The Output Data Products, usually much smaller than the input data, are transferred to the end user.

However, using most current standard facility infrastructures, the time required to download, process, and analyze the large volumes of remote sensing data is too long to effectively provide help in an emergency response situation. Cloud computing provides a novel way of quickly acquiring computer hardware and software for data processing only when they are needed.

Therefore, to support emergency response in crises (i.e., oil spills, tornados, flooding) in a time-effective manner, a cloud computing environment was used to process large sets of remotely sensed data. To accomplish and to expedite availability of remotely sensed data that would be necessary to address and respond to an emergency, software for processing time series of satellite remote sensing data was packaged together with a computer code that uses Web services so that the data sets from a NASA data archive and distribution system (LAADS) could be downloaded and utilized for emergency response purposes. The novel software package can be quickly deployed on a cloud computing platform only for as long as processing of the time series data is required to support emergency response. Multi-year time series of MODIS products can be created without the need for storing several terabytes of input data in a local system. Fast network connection between the cloud system and the data archive enables remote processing of the satellite data without the need for downloading the input data to a local computer system: only the output data products are transferred for further analysis.

Additionally, to further reduce operating costs, these computing resources can be released back to other users after the processing is completed. The processing software can be deployed on a cloud computing platform that provides a fast network connection to the NASA data archives. The data can then be downloaded just to the cloud computing system, without downloading the input datasets from the NASA data archives to the end user systems. Only the output data products, usually much smaller than the input data, are transferred to the end user for further analysis (see figure).

This packaged software for processing times series of satellite remote sensing data, combined with computer code that uses Web services to download the required data sets, enables the following: (1) quick deployment of MODIS time series production software on a cloud computing platform; (2) movement of times series production towards MODIS data archive to avoid transferring large input datasets to user’s system; (3) Web services to identify MODIS datasets for processing and to transfer the data from a data archive to the processing system; and (4) hardware and operating system to be quickly procured (via Web interface) only when required for the software deployment, and then released when no longer needed. This approach addresses historical barriers to technology transfer, where in-house resources to generate remote sensing products are not available or suitable for use in an emergency response situation. With this processing software that has been developed for cloud computing, data products can be produced in a timely manner, and then easily and economically deployed as needed.

This work was done by Bruce Spiering of Stennis Space Center and Slawomir Blonski of CSC. For more information, contact the Stennis Space Center Office of the Chief Technologist at 228-688-1929. Refer to SSC-00389.