Marine Geospatial Ecology Tools Announcements

The Marine Geospatial Ecology Tools Development Team would like to update you on recent MGET developments. You are receiving this message because you have subscribed to the mget-announce or mget-help email lists. If you do not wish to receive similar messages in the future, please visit the preceding links to remove yourself from the lists.

MGET demonstration at NOAA Coastal GeoTools 2011 conference

Next week, Ben Best and I will be demonstrating some of the latest tools added to MGET at the NOAA the Coastal GeoTools 2011 conference. The demo will be at the Tools Showcase session on Wednesday March 23 from 4:00-6:00 pm. Although the conference program describes specific demonstrations involving species diversity mapping and oceanographic processing, we will be flexible and can demo requested scenarios on demand. If you are attending GeoTools, please drop by our booth. We hope to see you there.

New tools for fitting and predicting multivariate tree models and for detecting eddies in Aviso SSH images

At long last, we are releasing some frequently-requested tools. In MGET 0.8a23 we added the Fit Tree Model and Predict Tree Model from Rasters tools. These tools fit multivariate classification and regression trees to tabular data and create predictive rasters from input rasters representing the value of predictor variables. They work just like the existing GLM and GAM modeling tools in MGET. An interesting feature of Fit Tree Model is that supports both automated and interactive tree pruning. This is the first tool we have developed that presents an interactive graphical user interface implemented in R. It works fairly well and we plan to implement similar capabilities in other tools in the future.

In MGET 0.8a25, to be released this weekend, we are adding the Find Okubo-Weiss Eddies in Aviso SSH Product tool. This tool downloads Aviso SSH images and executes the Okubo-Weiss eddy-detection algorithm and outputs a time series of rasters showing the locations of eddy cores. This tool was delayed for over a year due to compatibility problems between ArcGIS and the MATLAB Component Runtime that is used by the code that implements the algorithm. We have an adequate workaround to those problems now and are excited to finally release this tool.

As always, if you have any questions about MGET, please do not hesitate to contact us at or email me directly.

Best regards,

Jason