Practice and Example Datasets

To facilitate student participation, we provide several example geospatial datasets that can be used directly (including temperature, precipitation, and wildfire burn area data for the corresponding regions in 2024), or participants may use their own data and study areas.

Example Datasets

Blue Mountains, New South Wales (Australia)
A fire-prone mountainous region just west of Sydney, frequently affected by extreme bushfires such as the 2019–2020 Black Summer fires. Useful for studying peri-urban wildfire risk and climate influence.
data file
Alpine Region, Victoria (Australia)
Located in the Victorian Alps northeast of Melbourne, this high-risk bushfire area experiences repeated large-scale wildfires due to dry summers and rugged terrain.
data file
Butte County, California (USA) Site of the 2018 Camp Fire, the deadliest wildfire in California history. Suitable for studying extreme wildfire impacts and recovery. data file
Shasta County, California (USA) Known for recurring large wildfires, including the 2018 Carr Fire. Features varied topography and dense vegetation. data file
Boulder County, Colorado (USA) Located at the plains-mountain interface. High wildfire risk due to dry grasslands, suburban expansion, and strong wind events. data file

Participants are encouraged to explore these examples or bring their own datasets. GPBoost is especially effective in settings where structured covariates (e.g., reflectance bands, indices) and spatial dependence exist.

# Practice and Example Datasets {.unnumbered}

To facilitate student participation, we provide several example geospatial datasets that can be used directly (including temperature, precipitation, and wildfire burn area data for the corresponding regions in 2024), or participants may use their own data and study areas.

### Example Datasets

* **Blue Mountains, New South Wales (Australia)**  
  A fire-prone mountainous region just west of Sydney, frequently affected by extreme bushfires such as the 2019–2020 Black Summer fires. Useful for studying peri-urban wildfire risk and climate influence.  
  [data file](https://github.com/ausgis/igarss25ss/tree/main/data/student_challenge_datasets/sydney_100km)

* **Alpine Region, Victoria (Australia)**  
  Located in the Victorian Alps northeast of Melbourne, this high-risk bushfire area experiences repeated large-scale wildfires due to dry summers and rugged terrain.  
  [data file](https://github.com/ausgis/igarss25ss/tree/main/data/student_challenge_datasets/victoria_100km)

* **Butte County, California (USA)**
  Site of the 2018 Camp Fire, the deadliest wildfire in California history. Suitable for studying extreme wildfire impacts and recovery. [data file](https://github.com/ausgis/igarss25ss/tree/main/data/student_challenge_datasets/butte)

* **Shasta County, California (USA)**
  Known for recurring large wildfires, including the 2018 Carr Fire. Features varied topography and dense vegetation. [data file](https://github.com/ausgis/igarss25ss/tree/main/data/student_challenge_datasets/shasta)

* **Boulder County, Colorado (USA)**
  Located at the plains-mountain interface. High wildfire risk due to dry grasslands, suburban expansion, and strong wind events. [data file](https://github.com/ausgis/igarss25ss/tree/main/data/student_challenge_datasets/boulder)

Participants are encouraged to explore these examples or bring their own datasets. GPBoost is especially effective in settings where **structured covariates** (e.g., reflectance bands, indices) and **spatial dependence** exist.