EOS Volcanology Logo 1996 Progress Report

November 1996

1. Summary
2. Context of our Investigation
3. Data Validation Activities
4. Interactions with the DAACs
5. Contributions to the Educational Objectives of MTPE/EOS
6. Programmatic Contributions
7. 1996 Changes in Research Plans and Plans for FY97
8. List of Publications
9. Summary of our Most Important Accomplishments


This Progress Report describes both the scientific advances and programmatic issues related to the EOS Volcanology Interdisciplinary Team during 1996. We focus our descriptions on our preparations for science data products, and on our interactions with the DAACs and EOS Project. These activities include developing algorithms at the MODIS SCF for the routine nighttime detection of thermal anomalies using MODIS, our preparation for the placement of test data products at EDC and ASF DAACs, the development of routine TOMS volcano data products at Goddard DAAC, hosting an SO2 Validation Workshop at Goddard, and our participation in the EOS Q/A workshop (also at Goddard). The PI has also served as co-lead author for Chapter 10 of the EOS Science Paper (with Dennis Hartmann).

We have had two team meetings over the last 12 months where the specifics of our data product algorithms and their delivery schedules have been discussed. These algorithms have been formally reviewed by other investigators within our team. We now have a set of 10 data products for which we have formal algorithm descriptions. Descriptions of these data products are available on the Web at:


Our Team has also provided general programmatic support to the EOS project via our participation in the Q/A reviews run by R. Lutz. We have discussed HDF issues about data arrays and raster images with Brand Fortner (EOS-HDF developer) and Saud Amer (ECS representative at EDC), discussed ASTER data formats and Q/A with Moshe Pniel and Craig Leff at JPL, evaluated EOSDIS Evaluation Package 7, and worked with the ASTER Data Request Package (for Dale Noss). The PI for our Team also serves as the Chair of the Alaska SAR DAAC User Working Group.

In addition to these programmatic activities, our scientific results are described in 33 papers that we have published over the last 2 years (1995 - 1996), or have currently in press. These papers are listed in Section 8. Our published research includes obtaining quantitative data on the analysis of sulfur dioxide in volcanic eruption plumes detected by the TOMS and (as an ASTER analog) the TIMS instruments, studies of the decorrelation of radar images induced by new lava flows, and basic volcanological research into the atmospheric effects of volcanic gases. Many of our activities are also related to community outreach, both at a professional level and for the general public. These activities include the development of collaborative activities with the Alaska Volcano Observatory to use satellite images to detect volcanic eruption clouds.


This Interdisciplinary Science Investigation's objectives are four-fold:

1) To understand the physical processes associated with volcanic eruptions, including the eruption and cooling of lava flows, the injection of eruption plumes into the atmosphere, the derivation of digital topography for volcanoes (including the analysis of topographic change);

2) To investigate the manner by which ash, sulfur dioxide, water vapor, carbon dioxide, and other volcanic gases are injected into the troposphere and stratosphere, and the chemical processes associated with this release of volcanic gases into the atmosphere;

3) To place the diverse volcanic eruptions into the context of the regional tectonic setting of the volcano;

4) To educate volcanologists, school children and teachers, the general public, and our university peers about how remote sensing can be used to study volcanoes.

The closest match with EOS goals is in the understanding of the role of volcanism in climate change and volcanic inputs into the atmosphere. This is one of the seven high-priority areas of EOS research and we feel that our Team has taken the leading role within the EOS Project in articulating the importance of understanding eruption processes when investigating the influence of volcanism on climate. In addition, our investigation draws upon many of the Solid Earth objectives in hazard mitigation due to volcanic eruptions. Our Team has been instrumental in the development of algorithms for the near real-time analysis of transient phenomena (eruptions, wild fires), and in the derivation of digital topography using radar interferometry that will be of wide use to many Earth Science disciplines. Some of the ideas developed through our analysis of hazards has already proven helpful in providing outreach to the Hawaii State Civil Defense, who are in the process of setting up the Pacific Disaster Center on Maui and need technical input into how to use satellite remote sensing.


Arlin Krueger hosted a sulfur dioxide validation workshop last year to set up a framework for collaboration between ground and space observations, and to evaluate the sources of error for high altitude volcanic cloud retrievals. Other members of the EOS Volcanology Team who attended were Vince Realmuto, Ian Sprod and Bill Rose. The primary focus of this activity was the validation of sulfur dioxide, ozone, and ash/aerosol retrievals. We used EOS funding to set up a database containing ground observations of sulfur dioxide and ash optical depths in chance overpasses of volcanic clouds over the network of Brewer stations. This EOS funding was also used in partial support of an intercomparison of COSPEC instruments as modified for high altitude cloud validation and to develop software for comparison of TOMS data with TIMS and ASTER data.

We have had several field validation experiments for the SO2 and topographic algorithms. Much of this work was jointly sponsored by NASA's Geology Program, the SIR-C Project and Topography Program. The recent PacRim deployment of AES, TIMS and TOPSAR to Hawaii and New Zealand is an example of this use of funds from multiple programs to help validate retrievals of volcanic gas abundances, surface temperatures, and topographic data.


This last year we submitted a revised Product Data Plan to ECS. We also held our own 2-day algorithm developers' meeting at JPL in August 1996, which was attended by Saud Amer (EDC) and Nettie LaBelle-Hamner (ASF) so that our specific DAAC data issues could discussed. As Headquarters is aware, we have had numerous discussions with both EDC and ASF DAACs about where our Level 4 products will be archived. Until very recently (November 1996), our preference has been to have the majority of our products (excluding the TOMS products and some radar products) placed into EDC DAAC. Only by following this route could we be assured that the entire community will have access to these data, that the products will be preserved for the duration of the EOS Project, and that the metadata will exist for users to search the IMS to locate our products.

However, questions about funding the DAACs to work with our products is still being debated by Headquarters and the EOS Project Office. In October 1996 we developed a costs model for the two options in order to make our recommendations on how to proceed, and presented this model to Bryan Bailey at EDC. One model involved the placement of our Level 4 products within EDC DAAC. This work also included efforts by Howard Zebker to place one of our test data products (a radar-derived digital elevation model) into both DAACs to evaluate the process of writing custom HDF products and defining all of the metadata for a custom IMS search. Very recent communications with Bryan Bailey (November 20, 1996) indicate that other EOS Project priorities, specifically the need to ensure "at-launch readiness," means that IDS data products are not a priority item for the DAAC managers. EDC DAAC has not even formally requested that a cost model be prepared for their part of the activities.

As a result, our IDS Team decided at our November 1996 meeting that we would go with our second option for archiving our data products. While less that ideal, this option has a short-term benefit in terms of ease of implementation (and, hence, probably lower near-term costs) but has major draw backs in terms of longevity and search capabilities for the rest of the community. To ensure that our products are still visible to the outside world, we will continue to work with the DAACs to ensure that our metadata are included in the ECS IMS. We are about to start discussions within the Team about how we will address both the long-term archiving and distribution of our products; we also have a contact person from a commercial company that has considerable experience in handling very large planetary data sets that may be able to help our Team.


Our 1993-1995 Biannual Report outlined several outreach activities that we feel have established an excellent outreach program to our peers and to the community at large. These activities include:

1) Steve Self led the Decades Volcano Workshop at Taal in the Philippines in October 1995, where we presented many of the uses of remote sensing of Taal to the international volcanology community. Several of our students also attend this workshop and made good contacts with volcanologists at the Philippine Volcano Observatory. Some collaborations are developing here, and we are currently providing them with assistance on the interpretation of satellite images of Taal.

2) Bill Rose has helped established a network of agencies in Alaska that utilize satellite data sets to provide early warning of the hazards associated with volcanic eruption clouds. This network includes the Alaska Volcano Observatory (USGS), the University of Alaska, National Weather Service, the Federal Aviation Authority, and the State of Alaska. It is also worth noting that Dave Schneider, who is finishing his Ph.D. degree with Bill Rose, has taken a new job in Anchorage to work with the Alaska Volcano Observatory on the use of satellite data sets for volcano eruption plume detection for aircraft hazards. We are confident that having Dave in this position will greatly help the community become more proficient in the use of remote sensing data.

3) The EOS IDS Volcanology Team World Wide Web page has had 60,000 connections to its Home Page, and 16 times that many connections to its full set of Web pages. Additional volcanology Web sites that have a strong remote sensing and volcanology components are maintained by our Team Members, notably Bill Rose, Chuck Wood, and Arlin Krueger.

4) In addition to these ongoing activities, Pete Mouginis-Mark has recently increased his efforts to consolidate the MTPE and NASA Educational Affairs Office (Code FE) via the Space Grant Colleges. Space Grant has a national network of educators and researchers who are interested in MTPE activities, but who lack the detailed knowledge of the ongoing programs. Mouginis-Mark has agreed to chair the Space Grant MTPE Education Working Group in order to enhance their participation in the teaching of Earth remote sensing and climate change at the K to 12, college, and professional levels.


1) Luke Flynn has taken the lead in our Team's interactions with the MODIS Team and our analyses of the MODIS data stream. Over the last year, Luke has produced the software code for a Level 1B nighttime volcano alert, and this code was delivered to the MODIS SCF in July 1996. Volcanic eruptions will be partially discriminated from fires on the basis of spectral and spatial masks. This is an algorithm that will automatically search the MODIS Level 1b data stream at the MODIS SCF that will send an output file to the Hawaii SCF for each MODIS data granule (1354 x 2000 km). Geolocation will be done at Goddard SFC. The alert files will then be analyzed in Hawaii, and the distribution of thermal anomalies displayed on our Web site within a few hours of data acquisition by the spacecraft. This effort at the Hawaii SCF will perform the assessment of the alert files and will present the data on a continuously-updated global map. We are very excited about this data product, since it should be one of the first chances to see EOS data in near real-time.

2) The P.I. serves on ESSAAC (which is the main advisory group to the NASA MTPE Associate Administrator) and on the EOS Science Executive Committee (SEC). This work has also included contributing to the draft plan for the MTPE Plan for the Natural Hazards Program.

3) The P.I. and several team members contributed to "Volcanoes, Aerosols, and Climate Change," which is Chapter 10 of the EOS Science Implementation Plan.

4) Over the last year, both the Earth Probe TOMS and ADEOS TOMS have been launched. Arlin Krueger, as P.I. for the Total Ozone Mapping Spectrometers, will be delivering standard volcano SO2 data products to the Goddard DAAC.

5) Arlin Krueger has drafted a Memorandum of Understanding with NOAA/NESDIS for the near real time production of volcanic data from Earth Probe TOMS within the Alaska station mask. These products will be used to augment the AVHRR thermal alerts that NOAA, the USGS/AVO, and the University of Alaska are using to detect new volcanic activity along the Alaskan-Aleutian Arc.

6) The University of Hawaii has completed a Memorandum of Understanding with NASDA, the space agency of Japan, for the reception of JERS-1 and ADEOS data in Hawaii. We are also in discussions with RADARSAT International, EOSAT, and ESA for the reception of data from their spacecraft. These data streams will enable us to test many of the algorithms that we are developing for EOS, as well as provide the science community with data sets that would not otherwise be collected due to the limitations of direct broadcast satellite observations.

7) To help the radar community have a greater impact on the operations of the Alaska SAR Facility, the P.I. has served as Chair of the ASF DAAC User Working Group since May 1995.


Over the last year, we have re-evaluated our ability to use the MODIS data sets for daytime eruption detection. The issue here is that within the Level 1b data stream that we can search at the MODIS SCF, we are only able to interrogate 5 MODIS channels (Nos. 21, 22, 29, 31, and 32) in the time available. This means that we cannot search both the daytime and nighttime hemispheres (the bands we would use would be different). To generate the highest number of useful alerts using no more than 5 channels and only a few channel comparisons, we have opted for just a nighttime thermal alert rather than an S02 or daytime thermal alert.

Other product generation decisions have also been made on the basis of the funds and manpower that are available. We have decided that the low-resolution thermal map will now be ready only 1 to 2 years after launch. The nighttime thermal alert is consuming too much of our time to be able to deal with two MODIS data products at the same time given the people working this problem. Similarly, due to lack of funding, we have dropped our data product #3263, which was the eruption plume aerosol maps from MODIS data.

In support of the main activities listed above (Sections 2-4), we have identified the following key activities for our IDS investigation:

We will analyze active and potentially active volcanoes using interferometric radar techniques and data collected by existing sensors. A large part of this is to understand the distortions induced in the observed signatures due to atmospheric propagation. These signals must be quantified and accounted for in the data interpretation. The following subtasks will be performed:

a. Refine and complete in greater detail the three above mentioned algorithms.

b. Produce automated implementations of the algorithms.

c. Begin validation experiments to establish the algorithm performance in each case.

d. Refine the code to improve performance as suggested by the verification experiments.

e. Format metadata products according to EOSDIS standards and find a long-term archive for the radar products outside of the existing DAAC structure.

As a second aspect of our work for FY97, Luke Flynn will focus the MODIS thermal alert development on:

a. The format of the alert display on a short-term (one week) WWW page, and the planning for the long-term archive of the weekly summaries;

b. The development of the spatial mask for volcanoes (in order to facilitate the removal of many fires from areas where it is known that no volcanoes exist);

c. The development of the spectral mask to determine the intensity of the event and the difference between fires and eruptions. In terms of the software for the alert, updated versions have to be delivered to the MODIS SCF in January and July 1997. The final version of this code is due in January 1998, so that the development of this final version will be conducted during our upcoming funding period.

Several of our data products continue to evolve. With the new data stream from the EarthProbe and ADEOS TOMS instruments, our SO2 algorithm will be improved so that we will also recover the aerosol optical depth, which will essentially be a new product within the TOMS products. Similarly, Bill Rose and Gregg Bluth plan to look very closely at the range of geographic areas where the IR retrieval of ash loading in an eruption plume is taking place. We have seen signs in several AVHRR images that this retrieval does not work in a consistent manner when the background target is sea ice or snow. Further work using additional AVHRR scenes will be conducted to study this high latitude phenomenon.

We are also at a point where some of our software can be exported to other platforms. The derivation of the high resolution SO2 maps using ASTER data is one example. Currently, this code runs in IDL on a UNIX workstation. Vince Realmuto plans to port this code to an NT computer so that it will be more accessible to a wider community.

At our Team meeting in November 1996, we decided that in order to have most of our data products ready for distribution to the community at launch and on-budget, we will have to primarily rely on our own Web servers for their distribution. We will also continue our discussions with EDC and ASF DAACs regarding the metadata formats so that our products will be included in the IMS searches. However, we will investigate the possible use of commercial data archiving facilities to both ensure long-term stability of our products and off-load the network activity to mainland sites capable of providing fast Internet access (Hawaii currently has a band-width problem to the Mainland so this is not a good option, and we wish to avoid dedicated servers at our team member SCFs if at all possible due to cost considerations). "Tire-kicking" of the IMS and the other EOS data interfaces will also be required to ensure that the community can request their volcano- specific data sets.

In order to address the concerns of the reviewers of our 1993 - 1995 Progress Report, which said that our Team did not appear to lack cohesion, we have decided on the following activities:

a. We are planning two Team papers for the open literature that describe how our different data products will be used to study surface eruptions and plumes. We are hoping that data collected on the PacRim deployment may provide us with a real life opportunity to combine studies of TIMS, AES, and interferometric SAR for the eruption on Manam volcano in NE New Guinea. However, we have not yet seen these data, and so will make the final decision about the topic of study after that time.

b. We intend to organize a special session at the Fall 1997 AGU Meeting, that will showcase our latest science results as well as show the community the diversity of volcanology data products that EOS will provide.

c. We will have several Team members serve as "tire-kickers" of each others' software to make more of the Team familiar with each others data products. Here, the objective is to build synergism between the different data products so that we can show the community an "end-to-end" remote sensing study of a volcano rather than a single data product.




In Press:


Over the last year, our investigation has made significant progress in several areas of remote sensing techniques for the analysis of volcanoes and volcanic eruptions. In addition to the 33 papers that we have published or have in press, the highlights of our investigation include:

1) Our "Thermal Alert" code has been delivered to the MODIS SCF and accepted by them as part of the Level 2 MODIS production.

2) In conjunction with our NASA Topography Program efforts, we have modeled the radar interferograms to be expected from a dike intrusion on Mauna Loa. This is the first time that a time-series has been constructed for an intrusive event on a volcano. We will use this model as a comparison with real data to be collected in the future.

3) We had a series of press releases associated with our study of the Ruapehu eruption in New Zealand. These releases (issued by Fred Prata via CSIRO in Australia) made the Associated Press wire service, and appeared in many newspapers worldwide.

4) We have increased the robustness of our algorithms so that data from multiple satellites can now be used. This has been demonstrated by our use of ATSR data from ERS-1 to validate and extend our AVHRR observations of the 1994 eruption plume from Rabaul volcano. Using the ATSR data, we can confirm that a major component of the cloud was water ice, and that there were two distinct levels to the cloud (at about 11-13 km and at about 19-20 km). These stereo data provide us with an excellent analog to the MISR data sets that we hope will also be collected for large eruption plumes.

5) As part of the EP-TOMS mission, we have achieved the first detection of SO2 fumes from a non-erupting volcano. At Popocatepetl in Mexico, quiescent S02 degassing was detected by TOMS because of the lower spacecraft altitude and the resultant higher spatial resolution (25 km/pixel). This is very exciting, because it suggests that we can measure other degassing volcanoes (e.g., Etna and Erebus) in a comparable manner. The size of this detected anomaly, and the low altitude of the plume, provides us with greater confidence that we will be able to determine the details of volcanic S02 emissions with ASTER.

6) Several of our Team have participated in NASA Geology Program's PacRim deployment with the specific objective of collecting data to validate our algorithms and/or build baseline data sets for EOS. We participated in the collection of TOPSAR data over Hawaii, and the field validation of AES/TIMS data over White Island, New Zealand.

7) In conjunction with the NASA Space Grant program (Code FE), we have established a MTPE Working Group that will promote MTPE education at the K - 12 and college levels.

Back to EOS Volcanology Team Documents

EOS Volcanology LogoReturn to Main EOS Volcanology Page