Subject: Marine Geospatial Ecology Tools (MGET) help
Text archives
| From: | "Jason Roberts" <> |
|---|---|
| To: | "'Ben Best'" <>, "'Arthur Knight'" <> |
| Cc: | <> |
| Subject: | RE: [mget-help] Additional advice on MGET please |
| Date: | Fri, 4 Jun 2010 11:55:25 -0400 |
|
Hi Arthur, It sounds like you have some very specific ideas in mind for your
project. Discussing them with Ben, as he recommends, might be the quickest way
to determine if MGET will be of use to you. But based on your description, I
identified four areas that MGET might help you with: 1.
Converting data to ArcGIS-readable formats. MGET has various tools
for this. Whether they will be useful depends on what input data you will be
using. Many oceanographic datasets are distributed in HDF or NetCDF format.
ArcGIS has trouble with these and MGET can help you convert them to Arc format.
There are some examples here. 2.
Fitting statistical models and creating predictive maps from rasters
representing model covariates. MGET interoperates with R to fit GLMs and GAMs.
Once a model is fitted, you can evaluate its performance using ROC analysis and
then create maps of the predicted response. MGET does the work of reading your
rasters, running them though the prediction functions in R, and creating
outputs representing the response. There is a worked example here.
As Ben mentioned, we have not yet ported his CART implementation from ArcRstats
to MGET. It is likely that we will complete that job in the next few months, as
well as an implementation of RandomForests for evaluating the importance of
predictor variables in tree-based models. 3.
Incorporating R code into ArcGIS geoprocessing workflows. MGET
includes generic geoprocessing tools for executing
R code within geoprocessing models. If you have fragements of R code that
need to run at different steps in an ArcGIS workflow, this can be a convenient
way to orchestrate the flow, even if nothing else in MGET is useful to you. 4.
Efficiently building connectivity networks and calculating
least-cost paths between sites. These are the ConnMod tools that Ben has been
updating in preparation for incorporation into MGET. It is best if he explains the
details of them. I will also forward you a copy of the following paper which gives
an overview of MGET’s software architecture and an example of #2 above. Roberts, J.J., B.D. Best, D.C. Dunn, E.A. Treml, and P.N. Halpin
(2010). Marine Geospatial Ecology Tools: An integrated framework for ecological
geoprocessing with ArcGIS, Python, R, MATLAB, and C++. Environmental Modelling
& Software 25: 1197-1207 doi:10.1016/j.envsoft.2010.03.029. Best regards, Jason From: Ben Best
[mailto:] Hi Arthur, On Thu, Jun 3, 2010 at 8:43 PM, Arthur Knight <>
wrote: Hi. I am writing a review and framework for a new connectivity approach
to assessing wetlands in Queensland. I would like to reference some of your
software because of its use of CART and other statistical techniques, and
because the software may be applicable. However I have not heard back from Ben about accessing further
information and I could really do with that information now. Please let me know if you feel the MGET software is appropriate for the sort of analysis I am interested in (see
below). I would also be interested in hearing if you know of other
researchers who are using your software for wetland connectivity assessment and
management applications, as this would help to consolidate my decision on the
software. Regards Arthur Knight. ------------- Hi Ben. Do you have any advice for me about the most suitable software from
your suite of applications? I would appreciate your advice. Regards Arthur Knight. From: Arthur Knight Hi Ben. I have had a quick look at MGET and I thought it had a lot of tools
not relevant to my area of interest. My work requirement is to develop new, more consistent and
generally applicable assessment and mapping methods for hydrologic and ecologic
connectivity between terrestrial aquatic ecosystems for management
applications. This encompasses runoff-runon processes, defined flows and
rainfall but applying it across the rather large State of Queensland,
Australia, for all wetland types and geographic contexts. The techniques will be linked to existing mapping and inventory
data including DEMs and stream networks. A significant amount of the data will be stochastic in nature but
evaluated within a landscape model stratified by eco-hydrological sub-regions.
Given that the task is a multi-scaled spatio-temporal problem, I felt a more
defendable approach would be to include CART styled modelling to evaluate the
significance of connectivity characteristics for the inundation of different
wetland types with different environmental conditions and reference periods of
assessment. Scott Goetz and Patrick and Claire Jantz (Remote Sensing of
Environment 113 (2009) 1421-1429 published a terrestrial habitat graph theory
based connectivity study for NE USA using your ArcRstats package. This
application prompted my attention to your software. I am an SPLUS (similar to R) and ArcGIS user/programmer so I am
particularly interested in exploring how your software might be integrated into
our scientific and knowledge based systems. We also map and model estuarine and marine systems out to 3
nautical miles from the coast. Please let me know if you feel the MGET software is more
appropriate for evaluation. I would also be interested in hearing if you know of other
researchers who are using your software for wetland connectivity assessment and
management applications. Regards Arthur Knight. From: Ben Best [mailto:] Hi Arthur, On Thu, May 20, 2010 at 8:35 PM, Arthur Knight <> wrote: Hi. I
am interested in your ArcRstats Version 0.7. I have downloaded the zip file
(from your .duke.edu/geospatial/software site) but
my system tells me the zip file is corrupt, and it will not open. Could
you please confirm the status of the file for download? Thank
you. Arthur
Knight. ________________________________________________________ Dr
Arthur Knight Senior
Biodiversity Planning Officer Biodiversity
and Resource Management, Natural Resources and Environment Telephone 07
5451 2247 Facsimile 07
5451 2260 Email: Department
of Environment and Resource Management Centenary
Square, 52-64 Currie Street Nambour Q 4560 PO
Box 573, Nambour Q 4560 ________________________________________________________ Dr
Arthur Knight Senior
Biodiversity Planning Officer Biodiversity
and Resource Management, Natural Resources and Environment Telephone 07
5451 2247 Facsimile 07
5451 2260 Email: Department
of Environment and Resource Management Centenary
Square, 52-64 Currie Street Nambour Q 4560 PO
Box 573, Nambour Q 4560 +----------------------------------------------------------------+ Think
B4U Print 1 ream
of paper = 6% of a tree and 5.4kg CO2 in the atmosphere 3 sheets
of A4 paper = 1 litre of water +----------------------------------------------------------------+
|