Adj. Assist. Prof. Walt Meier's Research Students  


2001 - 2002 Academic Year

  1. Using Pathfinder satellite-derived SSTs to represent in situ data: Comparisons on the Florida Keys, Bahamas, and Great Barrier Reef, by Jonathan Shannon (co-advisors:  Dr. Marguerite Toscano and Dr. Kenneth Casey, NOAA) [Abstract]
  2. An analysis of Columbian coca cultivation via remote sensing, by Robert D. Holt [Abstract]
  3. An evaluation of the validity of an automated sea ice classification system for interpreting SAR imagery, by Sarah L. Heidt [Abstract]
  4. A Comparison of passive microwave algorithms for sea ice concentration, by Nathan Bastar [Abstract]
  5. Meier, W.N., S. Andersen, and N. Bastar, 2002.  Evaluation of SSM/I-derived sea ice concentrations with AVHRR imagery, Proceedings 2002 International Geoscience and Remote Sensing Symposium, 24-28 June, Toronto, ON.

2002 - 2003 Academic Year

  1. Comparison of Arctic sea ice motion from QuikScat and SSM/I, by Theodore Vermeychuk [Abstract]
  2. Meier, W.N., and T. Vermeychuk, 2003.  Sea ice motion products from microwave imagery, American Meteorological Society 7th Conf. on Polar Meteorology and Oceanography and Joint Symposium on High-Latitude Climate Variations, 12-16 May, Hyannis, MA.
  3. Evaluation of the accuracy of SSM/I-derived ice concentration products near the ice edge, by Meghan Poukish [Abstract]
  4. A new model:  Evaluating the Coupled Ocean Atmosphere Prediction System (COAMPS 3.0), by Lisa Morgan [Abstract]



 Using Pathfinder satellite-derived SSTs to represent in situ data:  Comparisons on the Florida Keys, Bahamas, and Great Barrier Reef

Midshipman:  Jonathan Shannon (Fall 2001)

Co-Advisors:  Marguerite Toscano and Kenneth Casey (NOAA)

Abstract:  In 1998, mass bleaching and mortality of coral reefs worldwide, related to anomalously high sea temperatures, highlighted the necessity of monitoring and understanding the effects of oceanic temperature changes on coral reefs and whether these changes are indicators of global climate change.  The NOAA/NASA AVHRR Oceans Pathfinder program has collected and calibrated 17 years (1985-2001) of global satellite sea surface temperature (SST) data.  Comparison of Pathfinder data to in situ data from three locations has shown how biases differ with geography and oceanography, time of satellite retrieval, and time of year.  Day/Night satellite returns possessed the greatest correlation and least bias.  Each region showed their own biases that were factors of oceanography, climate, and other factors.  To substitute Pathfinder SST data for in situ data, scientists need to recognize these factors and take them into account. [back to top]

 An analysis of Columbian coca cultivation via remote sensing

Midshipman:  Robert D. Holt (Fall 2001)

Abstract: Cocaine trafficking from Colombia is an issue that the United States military and drug-enforcement agencies deal with daily. They have had moderate success in reducing the supply of raw coca in the past; however, production-monitoring techniques must be modernized in order to combat the increasingly sophisticated cultivation methods of 21st century cartels.  Satellite remote sensing provides the means to improve these techniques.  Four conditions: lambda, cost, periodicity, and resolution are essential in assessing the usefulness of a particular satellite to the crop-monitoring project.  The best spectral band for distinguishing coca from other vegetation was near IR.A survey of satellites meeting those criteria was conducted, with the ETM+ of LANDSAT-7 and the SPOT-4 multispectral sensor outperforming a host of others.  Acquiring an image of a known coca field from coordinates given to the satellite company was not accomplished, resulting in the incompleteness of that phase.  The various satellites were analyzed using pre-existing and experimental algorithms to provide the best possible accuracy of the field.  Finally, the transfer of the database and image into a GIS format that the law enforcement agencies can utilize is discussed in depth.  This was done to improve the current techniques of 
mapping and analyzing the cultivation of coca in Colombia.  Integrating the improved satellite imagery into a GIS database will increase the accessibility and the ease of analysis of the data collected by the satellites.  [back to top]

 An evaluation of the validity of an automated sea ice classification system for interpreting SAR imagery

Midshipman:  Sarah L. Heidt (Fall 2001)

Co-Advisor:  Kyle Dedrick, U.S. National Ice Center

Abstract:  ARKTOS (Advanced Reasoning using Knowledge for Typing Of Sea ice) is a sea ice classification system that utilizes image processing and knowledge based rules to interpret RADARSAT SAR images, based on the Dempster Shafer Algorithm.  The system was created by the University of Kansas and is currently undergoing evaluation by the National Ice Center (NIC) and the Naval Research Laboratory (NRL).  The evaluation uses NIC charts, manually produced by ice analysts, as comparative data to determine the accuracy of ARKTOS products.  Although the initial attempt at making such a system, which can classify, open water, new ice, first year ice, and multiyear ice is impressive, there is still much to be done before ARKTOS can become operational.  This evaluation, although preliminary, indicates that there are many errors in the ARKTOS classification scheme and that the proficiency of ARKTOS appears to be limited in most circumstances.  Careful modification of its knowledge-based rules may be required on a case-by-case basis in order to improve concurrence between ARKTOS and NIC analyses.  [back to top]

 A Comparison of Passive Microwave Algorithms for Sea Ice Concentration

Midshipman:  Nathan Bastar (Spring 2002)

Abstract:  Accurate measurement of sea ice concentration is important for shipping and 
maritime travel in the Arctic and Antarctic Oceans.  Also, having a long-term record of sea ice concentration can aid scientists who study climate change and it’s effect on sea ice concentration.  Passive microwave imagery from the DMSP SSM/I sensor is the preferred method to study sea ice concentration because passive microwaves can be detected through clouds and most weather conditions.  The brightness temperatures obtained from different microwave frequencies are incorporated into an algorithm that determines the concentrations.  In this study six different algorithms - Bristol, Bootstrap, Cal/Val, NASA Team 2, NIC Hybrid and NASA Team - are compared with AVHRR imagery from June to early July 2001 and late December through January 2002.This data was taken over 3 different regions:  the East Greenland Sea, the Barents Sea and Baffin Bay.  This study found that there are differences in the performance of the different algorithms at the different times of year and in the different seasons.  The NASA Team and Bootstrap algorithms consistently produced the results with the smallest errors.  The Bristol works well in the winter and the NASA Team 2 works well in the summer.  [back to top


 Comparison of Arctic sea ice motion from QuikScat and SSM/I

Midshipman:  Theodore Vermeychuk (Fall 2002, Spring 2003)

Abstract:  Arctic sea ice motions for January through March 2002 derived from QuikSCAT and SSM/I data are generally well correlated over the Arctic Ocean as a whole.  Special sensor microwave/imager (SSM/I), passive microwave data, has been available for several years, but QuikSCAT, which uses active microwave and measures backscatter, is relatively new and few in-depth studies have been done.  Although both data sources utilize microwave imaging to measure the same parameter, preliminary observations suggested that each data set is independent.  More thorough tests indicate that the two data sources are not statistically independent, however each has different strengths, biases and errors.  Examining both data sets yields a more thorough picture of Arctic sea ice motion than either satellite alone.  Also, though the data examined is limited to points where data for both satellites is available (for correlation purposes), in some cases one of the satellites may provide better coverage for some particular areas than the other, although this study did not examine this possibility.  It remains to develop an algorithm for fusing the two data sets, or some other appropriate method of using both in conjunction for sea ice motion modeling purposes.  [back to top]

 Evaluation of the accuracy of SSM/I-derived ice concentration products near the ice edge

Midshipman:  Meghan Poukish

Abstract:  Among other remote sensing sources, composite imagery from SSM/I swaths is relied on daily by ice analysts to predict the ice edge in the Polar Regions.  It is imperative for maritime operations in those regions for their predictions to be as accurate as possible, but many sources of error exist.  Imagery from SSM/I sensors near the ice edge is inaccurate primarily due to resolution, temporal averaging, and weather effects.  There is also potential error in the transfer of data from individual swaths to composite imagery.  Variables such as weather and location affect the degree of this error.  This project will analyze the inaccuracies of the composite imagery from SSM/I sensors caused by these primary reasons as compared to averaged satellite swaths and individual swaths and consider solutions for improving accuracy.  Possible solutions include subtracting weather effects from the raw data to obtain atmospheric emission-free values and changing averaging procedures for the creation of composites. [back to top]

 A new model:  Evaluating the Coupled Ocean Atmosphere Prediction System (COAMPS 3.0)

Midshipman:  Lisa Morgan

Abstract:  The computer-based model presently employed by the United States Navy for meteorological forecasting is the Coupled Ocean Atmosphere Prediction System, or COAMPS 3.0.  This model is simple, and based is principally upon primitive equations including non-hydrostatic effects.  Unlike the even more basic Navy Ocean Global Atmospheric Prediction System (NOGAPS), it does, however, take into account interactions between ocean, air, and terrestrial components; thus it does not assume a uniform ocean void of landmasses, as did its predecessor.  Variables taken into account include wind components, potential temperature, mixing ratio, surface pressure, ground temperature, ground wetness, and sea-surface temperature.  The objective of this study is to expand the categorical statistics of COAMPS to include heavy rainfall threshold, temperature, surface pressure variations, and gale force winds.  This is done by conducting isolated satellite observations over the area bounded by 30? to 40?N Latitude and 60? to 70? W Longitude in the Northwest Atlantic Ocean.  Observations are conducted via the use of SSM/I and DMSP visible and infrared data collected from the Fleet Numerical Meteorology and Oceanography Center in Monterey, California.  These images are examined using ImageJ (Image Processing and Analysis in Java) software, and statistical verification is conducted to determine the effects of these parameters on weather modeling accuracy out to 72 hours.  [back to top]

Back to top