There will be another spatial-statistical-network (SSN) model training workshop in Boise, Idaho this fall October 27–29 and that registration is now open (register here: https://www.eventbrite.com/e/spatial-modeling-workshop-oct-27-29-2025-tickets-1474764533419?aff=oddtdtcreator).
If you are interested in attending (or have a student that is), they recommend registering ASAP because it usually fills rapidly once registration opens. If you’d like to come to the workshop but the financial arrangements with your employer will take time to set up, please contact Dan Isaak, daniel.isaak@usda.gov, and they’ll try to hold a spot for you.
Spatial-Stream-Network Models training workshop background:
A class of spatial statistical network model for data on stream networks has recently been developed & free software is available for implementing the models. SSNs account for network topology (i.e., flow direction, stream size, tributary confluences) and offer significant improvements over many traditional statistical techniques that were developed originally for terrestrial applications. The SSNs are applicable to common types of stream data (e.g., water quality attributes, biological survey information, genetic metrics, habitat conditions) through application of several distributions (e.g., Gaussian, binomial, Poisson). The models also account for spatial autocorrelation among measurements, which makes them powerful tools for mining information from large datasets aggregated from multiple sources or sampling designs that are clustered and nonrandom. For additional details, please visit the SSN/STARS (https://research.fs.usda.gov/rmrs/projects/spatial-stream-network) and National Stream Internet websites (https://research.fs.usda.gov/rmrs/projects/national-stream-internet).
The 3-day workshop consists of a 1-day short-course & 2 days of working with course instructors to apply SSNs to participant’s datasets. Note that attendance is limited to 20 students to ensure a high-quality experience and sufficient interaction with course instructors. Workshop registration is $700 for students/postdocs and $1,000 for professionals, which does not include lodging or meals.
The workshop does include:
-Sharing of free software tools used to conduct SSN analyses (SSNbler for spatial editing with GIS software (https://pet221.github.io/SSNbler/articles/introduction.html#background) and the SSN2 package (from GitHub: https://usepa.github.io/SSN2/ or CRAN:https://cran.r-project.org/web/packages/SSN2/index.html) for analysis in R)
-Demonstration of data processing steps necessary to calculate the spatial information needed to fit SSN models
-Demonstration of how the SSN2 package is used for:
- Exploratory data analysis
- Modelling continuous, presence/absence, and count data using spatial linear, linear mixed-effects, and generalized linear models
- Model diagnostics and selection
- Prediction (kriging) and block kriging (estimating totals for discrete areas)
- Uncertainty estimation
- Simulation and spatial visualization techniques for streams data using ggplot2
-Discussions of when spatial statistical techniques are most useful and the many new applications that are possible for stream data;
-Working with instructors to apply the SSNs to your data
*Note that a good working knowledge of statistics and the R statistical program are useful to obtain the greatest benefits from the workshop. Participants at the 3-day workshop must bring a laptop with access to R statistical software and a GIS of your choice (Recommended: ArcGIS Pro with standard or advanced license or QGIS software). For more background on SSN model theory, applications, and software developments, up-to-date bibliographies are maintained at the SSN & NSI websites linked above, as are example datasets and documentation associated with implementing SSN analyses.