DMS Node Focus

The University of South Carolina is not exactly a node of IRIS. However, we do receive some funding from the DMS to lead the development of the IRIS FISSURES project. This qualifies us as a "nodelette," but we've been granted a full node focus! DMS funding to USC supports only a quarter of programmer Philip Crotwell's time and one grad student. The project is coordinated by Tom Owens, who receives no DMS support. Our responsibilities include interacting with 2AB, Inc, the consulting firm that has guided us through the analysis and design process, advising groups wishing to make use of the FISSURES framework, and developing code for various elements of the Data Handling Infrastructure project.

FISSURES Design

The analysis and design on the FISSURES framework has depended heavily on the talents of Mitchel Sanders of 2AB, Inc. His extensive experience in designing CORBA based systems allowed a much better framework to be developed than would have been possible without his "adult supervision." We continue to interact with him in support of IRIS's contract with 2AB as the domain experts. This involves meeting face to face once every two to three months as well as reviewing revisions to the design to make sure they make sense in seismological terms as well as in software terms and preparation before each meeting. A large part of validating designs involves creating trial implementations for the various services. There are two main ways that these trial implementations are seeing the light of day, Data Handling Infrastructure (DHI) and South Carolina Earth Physics Project (SCEPP), our independently-funded high school seismology project.

Data Handling Infrastructure

The major IRIS-DMS project that utilizes FISSURES is the Data Handling Infrastructure (DHI) project. We have been advising the software developers at the DMC on implementing services that interact with the database and data storage system to deliver event and channel meta-data as well as waveforms via the FISSURES interfaces. This subset of the framework has been named the Data Handling Infrastructure and aims to more efficiently deliver data products to seismologists.

Efficiency can mean many things, from the simplest "how fast can the bytes move" to more important, "how much faster can the seismologist do his/her work." We are of the opinion that the first, while easy to measure, is too simplistic a metric and that the second is, in the end, the only one that matters. For example, while keeping all response information on the seismologist's local computer is much faster, the extra time spent retrieving this information, keeping it current with updates at the DMC and general management of the information greatly outweighs the small speed savings over letting the DMC handle the management and using a remote access mechanism to the database at the DMC.

The current efforts, which are underway, are to implement three FISSURES services. The first allows an application to query event information stored in the database. This will go a long way to addressing the limitations of WEED, for example, in allowing up to the minute events to be queried. The second is a network meta-data browser that allows network, channel and response information to be queried. The last is a waveform retrieval service that queries the new FARM and allows data to be pushed back in small pieces as they are extracted at the DMC instead of all in one huge SEED volume.

South Carolina Earth Physics Project

SCEPP is a high school seismology project at the University of South Carolina and uses the FISSURES framework as the basis for its data movement system. It is funded by the State of South Carolina and the University of South Carolina, not by IRIS . Thus, a purist might not include this activity in a DMS node focus. But, as a nodelette, we have to fill our space with something! Seriously, SCEPP brings several key ingredients to the FISSURES/DHI project: Motivation and Resources. SCEPP is the primary reason that we are investing our own resources (the bulk of Crotwell's salary) into the FISSURES/DHI infrastructure. We have a practical problem to solve! SCEPP serves as a useful testbed for FISSURES and DHI. We have a system built that recovers data from a A to D board, buffers it into a local database on a Linux PC at each high school, ships it upstream to our main server at USC and stores it in our permanent database. Under SCEPP, we have also developed a data explorer for the K-12 community called the Carolina Earthquake Explorer. This is a java-based application that allows non-seismologists relatively easy access to the DMC FARM and includes several learning modules intended to help teachers convey important concepts in Earth and Physical Science using seismic data. In fact, IRIS E&O seems to be leaning toward investing resources into a project to upgrade the Carolina Earthquake Explorer to fulfill their goal of providing an IRIS Virtual Seismic Network Explorer to the K-12 community. This provides a natural marriage of the next generation of DMC data access methodology and E&O's desire to take the lead in providing seemless data access to the K-12 community.

Submitted by Philip Crotwell and Tom Owens

For more information or comments contact