Fort Collins Science Center

Resource for Advanced Modeling

RAM Home

As of September 2009, the USGS Fort Collins Science Center has a newly renovated facility known as the Resource for Advanced Modeling (RAM), which provides a collaborative working environment for up to 20 scientists. This space is a facility for collaborative research within the USGS and the wider research community. There are networked, wireless computing facilities with the ability to run and test various models (e.g., Maxent, Boosted Regression Trees, Logistic Regression, MARS, Random Forest) for a variety of spatial scales (county, state, region, nation, or global). These techniques use predictor layers from MODIS time-series data as well as current and future climate layers (near- and long-term projections). The main purpose of the RAM is to bring together remote sensing and climate forecasting experts, habitat modelers, field ecologists, and land managers in a synergistic environment.

Photo of the main modeling area Photo of docking stations. Photo of the break out area.

The RAM has WebEx capability, podium, and projector. This workspace also features electronic team boards with screen-capture capability, white boards, and wireless internet capabilities. The collaborative workroom has the capability to quickly alternate between 6 laptops shown through the projector. It offers 3 docking stations in a separate work area, a small lounge for break-out discussions, a DVD/VCR unit, speakers, and a conference phone.

Directions to the RAM

The Resource for Advanced Modeling is located within the USGS Fort Collins Science Center in Fort Collins, CO. This facility is a quick 70-mile drive north from the Denver International Airport. For maps and detailed directions, see Driving Directions.

Contact

Tracy Holcombe
U.S. Geological Survey
Fort Collins Science Center
2150 Centre Ave., Building C
Fort Collins, CO 80526-8118
Tel. 970.226.9380
Fax 970.226.9230
holcombet@usgs.gov

 

Selected Model Examples
Datasets and Modeling Techniques available for Collaborators

Datasets

Data Set Extent Resolutions Years Layers
Daymet Continental US 1 km2 Monthly average of 1980-1997 Precipitation, frost days, frequency of precipitation, growing degree days, humidity, radiation, and temperature (max, min, and mean)
World Clim Global 30 sec, 2.5 min Average from 1950-2000 Temperature (min, max, and mean) and precipitation
World Clim Global 30 sec, 2.5 min 2020, 2050, 2080 Climate models (HADcm3, CSIRO, CCCMA), emission scenarios (A2a, B2b)
PRISM Continental US 800 m2 Average of 1971-2000 Temperature (min, max, and mean) and precipitation
VEMAP Continental US 0.5 degree 2099 Temperature (min, max, and mean) and precipitation for Hadley and CCC scenarios
USGS 1k Continental US 30 sec (~1 km2) n/a Elevation, slope, northness, eastness, compound topographic index, flow accumulation, flow direction
MODIS Vegetation Continuous Field (VCF) North America, South America, Oceania 500 m2, 1 km2, 4 km2 2001 Bare ground, tree cover, herbaceous cover
MODIS Phenology North America 250 m2, 500 m2, 1 km2 2001-2007 (except 2005), and average across years Beginning of season, end of season, length of season, base value, peak time, amplitude, greenup rate, browndown rate, integral over season (absolute), integral over season (scaled), maximum value, minimum value, RMSE

Modeling techniques

Model Technique

Input

Description

Reference

Logistic Regression Binary Presence/absence data prediction of habitat suitability McCullagh and Nedler 1989
Multiple Linear Regression Continuous Continuous data predictions (e.g., percent cover) using regression principles Most statistical packages
Maximum Entropy Modeling (Maxent) Presence Inferences from available data, avoiding unfounded constraints from the unknown (principles of maximum entropy) Phillips et al. 2006
Genetic Algorithm for Rule-set Production (GARP) Presence Successive iterations of a rule-set, modified each time, to convergence Stockwell and Peters 1999
Classification and Regression Trees (CART) Binary/ Continuous Repetitively partitions the dependent data into two homogenous groups at a node using regression principles Breiman et al. 1984
Boosted Regression Trees (BRT) Binary/ Continuous Constructs an "ensemble" of regression trees (CART) Elith et al. 2008
Environmental Envelope Presence Environmental range where present applied to other locations. Jarnevich et al. In review
Random Forest Binary A machine learning technique that generates many classification trees and aggregates the results Breiman 2001; Prasad et al. 2006
Multivariate Adaptive Regression Splines (MARS) Binary Regression model allowing non-linear responses Leathwick et al. 2006
Manually derived Presence Species-specific model generated by experts, implemented spatially in ArcGIS For example, see Rodda et al. 2009

Breiman, L. 2001. Random forests. Machine Learning 45: 5-32.

Breiman, L., J.H. Friedman, R.A. Olshen, and C.G. Stone. 1984. Classification and regression trees. Wadsworth International Group, Belmont, California, USA.

Elith, J., J.R. Leathwick, and T. Hastie. 2008. A working guide to boosted regression trees. Journal of Animal Ecology 77: 802-813.

Jarnevich, C.S., T.R. Holcombe, D.T. Barnett, T.J. Stohlgren, and J.T. Kartesz. [In review.] Forecasting weed distributions using climate data: A GIS early warning tool. Invasive Plant Science and Management.

Leathwick, J.R., J. Elith, and T. Hastie. 2006. Comparative performance of generalized additive models and multivariate adaptive regression splines for statistical modelling of species distributions. Ecological Modelling 199: 188-196.

McCullagh, P. and J.A. Nelder. 1989. Generalized linear models, 2nd edition. London and New York: Chapman and Hall.

Phillips, S.J., R.P. Anderson, and R.E. Schapire. 2006. Maximum entropy modeling of species geographic distributions. Ecological Modelling 190: 231-259.

Prasad, A.M., L.R. Iverson, and A. Liaw. 2006. Newer classification and regression tree techniques: Bagging and random forests for ecological prediction. Ecosystems 9: 181-199.

Rodda, G.H., C.S. Jarnevich, and R.N. Reed. 2009. What parts of the US mainland are climatically suitable for invasive alien pythons spreading from Everglades National Park? Biological Invasions 11: 241-252.

Stockwell, D. and D. Peters. 1999. The GARP modelling system: Problems and solutions to automated spatial prediction. International Journal of Geographical Information Science 13: 143-158.

Software for Automated Habitat Modeling (SAHM)

The RAM is equipped with a modeling program called the Software for Automated Habitat Modeling (SAHM). SAHM is used to create habitat models for endangered species and invasive species. The code provides a significant capability to run, in a consistent and repeatable manner, five different habitat models (Maxent, Boosted Regression Trees, Logistic Regression, Multivariate Adaptive Regression Splines, and Random Forest) and produces an ensemble result from all model techniques selected for a single run. An example of this ensemble result and a short discussion of each modeling technique is available in the paper by Stohlgren et al., linked below.

SAHM combines predictor layers (environmental raster data layers of the study area, such as those available in the RAM) with user-collected field sampling measurements for a particular species. The program uses these data to run statistical models that analyze habitat requirements of a species of interest and predict the potential distribution based on habitat suitability. Model outputs can help users, such as land and natural resource managers, generate predictive maps or reports to aid in predicting and managing the spread of invasive species.


System Requirements

  • The Java Run-time Engine (JRE) must be installed. 
  • The R modeling engine must be installed with the packages gbm, mda, PresenceAbsence, rgdal, sp, randomForest, and XML. The base R installation and these packages are available from the R web site, http://www.r‐project.org/, all of which may be installed according to the R documentation.

Contact

Catherine Jarnevich Jeff Morisette Tracy Holcombe
U.S. Geological Survey
Fort Collins Science Center
2150 Centre Ave., Building C
Fort Collins, CO 80526-8118
Tel. 970.226.9439
Fax 970.226.9230
U.S. Geological Survey
Fort Collins Science Center
2150 Centre Ave., Building C
Fort Collins, CO 80526-8118
Tel. 970.226.9144
Fax 970.226.9230
U.S. Geological Survey
Fort Collins Science Center
2150 Centre Ave., Building C
Fort Collins, CO 80526-8118
Tel. 970.226.9380
Fax 970.226.9230

Also at FORT

References for models that have come from the Resource for Advanced Modeling (RAM)

Carter, G.A., K.L. Lucas, G.A. Blossom, C.L. Lassitter, D.M. Holiday, D.S. Mooneyhan, D.R. Fastring, T.R. Holcombe, and J.A. Griffith. 2009. Remote sensing and mapping of tamarisk along the Colorado River, USA: A comparative use of summer-acquired hyperion, Thematic Mapper and QuickBird Data. Remote Sensing 1(3): 318-329.

Evangelista, P.H., S. Kumar, T.J. Stohlgren, C.S. Jarnevich, A.W. Crall, J.B. Norman III, and D.T. Barnett. 2008. Modelling invasion for a habitat generalist and a specialist plant species. Diversity and Distributions 14(5): 808-817.

Evangelista, P.H., T.J. Stohlgren, J.T. Morisette and S. Kumar. 2009. Mapping invasive tamarisk (Tamarix): A comparison of single-scene and time-series analyses of remotely sensed data. Remote Sensing 1(3): 519-533.

Holcombe, T., T.J. Stohlgren, and C. Jarnevich. 2007. Invasive species management and research using GIS. In: G.W. Witmer, W.C. Pitt, and K.A. Fagerstone (eds.). Managing vertebrate invasive species: Proceedings of an international symposium. Fort Collins, CO: National Wildlife Research Center. p. 108-114.

Holcombe, T.R. 2009. Early detection and rapid assessment of invasive organisms under global climate change. Ph.D. dissertation. Fort Collins, CO: Colorado State University. 112 p.

Jarnevich, C.S. and T.J. Stohlgren. 2009. Near term climate projections for invasive species distributions. Biological Invasions 11(6): 1373-1379.

Kumar, S., S.E. Simonson, and T.J. Stohlgren. 2009. Effects of spatial heterogeneity on butterfly species richness in Rocky Mountain National Park, CO, USA. Biodiversity and Conservation 18(3): 739-763.

Kumar, S., S.A. Spaulding, T.J. Stohlgren, K. Hermann, T. Schmidt, and L. Bahls. 2009. Potential habitat distribution for the freshwater diatom Didymosphenia geminata in the continental US. Frontiers in Ecology and Environment 7(8): 415-420.

Kumar, S. and T.J. Stohlgren. 2009. Maxent modeling for predicting suitable habitat for threatened and endangered tree Canacomyrica monticola in New Caledonia. Journal of Ecology and the Natural Environment 1(4): 94-98.

Li, M-Y., Y-W. Ju, S. Kumar, and T.J. Stohlgren. 2008. Modeling potential habitats for alien species of Dreissena polymorpha in the continental USA. Acta Ecologica Sinica 28(9): 4253-4258.

Morisette, J.T., C.S. Jarnevich, A. Ullah, W. Cai, J.A. Pedelty, J.E. Gentle, T.J. Stohlgren, and J.L. Schnase. 2006. A tamarisk habitat suitability map for the continental United States. Frontiers in Ecology and the Environment 4(1): 11-17.

Morisette, J.T., A.D. Richardson, A.K. Knapp, J.I. Fisher, E.A. Graham, J. Abatzoglou, B.E. Wilson, D.D. Breshears, G.M. Henebry, J.M. Hanes, and L. Liang. 2009. Tracking the rhythm of the seasons in the face of global change: Phenological research in the 21st century. Frontiers in Ecology and the Environment 7(5): 253-260.

Rodda, G.H., C.S. Jarnevich, and R.N. Reed. 2009. What parts of the US mainland are climatically suitable for invasive alien pythons spreading from Everglades National Park? Biological Invasions 11(2): 241-252.

Sutton, J.R., T.J. Stohlgren, and K.G. Beck. 2007. Predicting yellow toadflax infestations in the Flat Tops Wilderness of Colorado. Biological Invasions 9(7): 783-793.

 

Skip About Section navigation and continue to content

RAM Schedule

Top of Page