Schedule

Below is the schedule for GSP 470/570. We'll also be using content from the R For Spatial Stats web site.

Data sets:

• Example tree data from the FIA database
  • Contains CSVs by species with FIA plots that contain the target species. Attributes include Lat, Lon, Max/Min Diameter, Max/Min Height, and Count
• BioClim layers
  • CONUS at 1km - Standard layers cropped to Continental US.
  • Subtset of layers for North American at 4km
• Example files for Douglas-Fir trees in California
  • Count data for California - All FIA plots in California with the number of doug-fir trees
  • Presence data for California - Just FIA plots in California with one or more doug-fir trees
  • Maximum Height in FIA plots for all of California at 8km - Grid of points for all of California with the maximum height of doug-fir in each 8km grid cell
• Redwood trees in Humboldt County - I just recreated this data from the FIA database. It should contain counts and other information for redwood trees in FIA plots in Humboldt County.
• Dominant Trees for California
• Ref_Species Table from FIA - includes species codes (spcd) and Common Names

New resources:

• Creating Reports in MS-Word
• Report rubric
• Modeling Method Summary

1. Introduction to Geospatial Modeling

• Presentation: Home page and syllabus review (see Canvas)
• Presentation: Introductions
• Learning Module: Linear Regression
• Optional:
  • Plant, Chapter 1
  • Compute p-values
  • Check out this web site with dynamic stats tools
• For fun: Spurious Correlations

2. Synthetic Data

• Note that we will be using R Studio in lab. I highly recommend it over other environments.
• Presentation: Overview of Spatial Modeling
• Learning Module: Synthetic Data with R
• Optional Reading:
  • Plant, Chapter 2: R Programming Environment up to 2.3.2.
  • Synthetic data generation — a must-have skill for new data scientists

3. Nature of Spatial Data

• Learning Module: Nature Of Spatial Data
• Optional Reading:
  • Plant, Chapters 3 and 4 (read once but don't feel like you need to fully understand it all)

4. Generalized Linear Methods

• Optional Reading: Plant, Chapter 9, all of sections 9.1, 9.2, and 9.4.1
• Learning Module: GLMs
• Presentations:
  • Building Models
  • Creating Predictive Maps for Continuous Variables and Point/CSV Files

5. Generalized Additive Methods

• Reading: See learning module
• Presentation: Setting up for Modeling
• Learning Module: GAMs
• Reading for Thursday: Jensen, Winter distribution of blue crab Callinectes sapidus in Chesapeake Bay: application and cross-validation of a two-stage generalized additive model
  • Slides for discussion
• Reading (skim): Generalized Additive Models by Hastie and Tibshirani

6. Regression Trees

• Optional Reading: Plant, Chapter 8 (focus on 8.3)
• Converting Data Types - in case you need to convert your data to categorical
• Reading for Tuesday: Martínez-Rincón, Comparative performance of generalized additive models and boosted regression trees for statistical modeling of incidental catch of wahoo (Acanthocybium solandri) in the Mexican tuna purse-seine fishery
  • Figures from the reading
• Learning Module: CART

7. Estimating Uncertainty

• Optional Reading: Plant, Chapters 5, 6, and 7 (focus on 5.2.2 and all of 6)
• Presentation: Modeling Process
• Learning Module: Estimating Uncertainty
• Activity: Statistics without the pain

8. Habitat Suitability Modeling

• Learning Module: Habitat Suitability Modeling with HEMI 2

9. Habitat Suitability Modeling With MaxEnt

• Learning Module: Habitat Suitability Modeling with MaxEnt
• Readings and discussion: See Canvas

10. Agent-Based Spatially Explicit Models

• Learning Module: Agent Based Spatially Explicit Models
• Reading and discussion: See Canvas

11. Mixed Models

• Learning Module: Mixed Models
• Reading and discussion: See Canvas
• Lab: Work on projects

12. Bayesian

• Topic: Bayes
  • Optional reading: Probabilistic rainfall thresholds for landslide occurrence using a Bayesian approach
  • Interesting web page on diagrams for hierarchical models
  • Optional reading: Frequentism and Bayesianism: A Practical Introduction
• Lab: Work on projects

13. Other Topics

• Topic: Autocorrelation and Interpolation
• Topic: Modeling Summary
• Topic: Big Rasters
• Topic: HPC
• Lab: Work on projects

14. Going Global

• Presentation & Discussion: See Canvas
• Older optional stuff:
  • Presentation: How Western Science is Leaving Room for Faith in How the Universe Works
  • Old: What does the future look like?
• Reports: Humboldt Bay SLR Vulnerability and Adaptation Planning
• Maps: A Terrifying, Fascinating Time-lapse of 30 Years of Human Impact on Earth
• Wrap up on Modeling, Discussion on what we can do.
• Reading: How fossil fuels will save the world
• Writing Readable R Code
• Lab: Work on projects

Additional Resources

Converting Data Types

Applied Spatial Statistics In R

R Labs for Vegetation Ecologists from MSU

Regression Modeling Steps (incomplete)

RStoolbox: Tools for Remote Sensing Data Analysis

Additional Potential Topics

• Learning Module: Modeling Uncertainty in Models from MaxEnt
• Reading: Mapping sources, sinks, and connectivity using a simulation model of northern spotted owls
• Optional Reading:
  • Revised Recovery Plan for the Northern Spotted Owl
  • Endangered and Threatened Wildlife and Plants; Designation of Revised Critical Habitat for the Northern Spotted Owl
  • MODELING SUPPLEMENT to the Final Rule to Revise Critical Habitat for the Northern Spotted Owl
• Topic: Big Rasters
• Topic: Machine Learning Approaches
• Topic: High-Performance Computers
• Reading: The Fourth Paradigm, Forward and Introduction by Jim Gray
• Reading: A working guide to boosted regression trees
• Lecture: Discussion on The Fourth Paradigm
• Guest Lecture: Heather Lintz - Non parametric Multiplicative Regression
• Guest Lecture: Heather Lintz - Quantifying ecological thresholds from response surfaces

Old Learning Module: Obtaining and Managing Data

Reading: Graduate Student’s Guide to Necessary Skills for Nonacademic Conservation Careers

• Ask yourself, are you prepared for the career you are entering into? If not, what can you do to gain the knowledge, skills, and experience you need?

Old Lab: Synthetic Data and Polynomial Trend Surfaces with Excel and ArcMap