Data from: A high-altitude thermal infrared method for estimating moose abundance and demography in Rocky Mountain National Park, USA
Data files
Mar 17, 2026 version files 582.85 KB
-
2019_complete_data.csv
10.79 KB
-
2019_metadata.csv
1.43 KB
-
2020_complete_data.csv
11.32 KB
-
2020_metadata.csv
1.40 KB
-
hsi.zip
518.70 KB
-
MooseAnalysis2020_revised.Rmd
19.93 KB
-
README.md
3.64 KB
-
ROMO_Moose_Elk_IR_Survey_Strata.csv
1.06 KB
-
sample_grid.zip
14.56 KB
Abstract
Resource managers require accurate estimates of large herbivore abundance and demography to maintain ecological integrity. Common methods to count these species, including observations from low-altitude helicopter flights, may conflict with other protected area management objectives and struggle to produce precise estimates for more cryptic species. To address these issues, we evaluated thermal infrared (TIR) assisted counts of Shiras moose Alces alces shirasi in a temperate montane ecosystem of Colorado, USA. We binned the 694.7 km2 study area into two strata, depicting higher and lower moose habitat suitability. Sixty transects (each 536 m wide by 5 km long) were randomly selected and flown from fixed-wing aircraft flying at altitudes centered around ~ 610 m (2000 ft) above ground level (AGL) in July 2019 and again in July 2020. We determined abundance and demographic information using double-observer line distance sampling with estimates produced using n-mixture models. The detection probability of groups on the line, and of individuals within a group, conditional on the group being observed, was high. Mean moose density estimates across the study area were consistent from year to year, with reasonable confidence. We estimated 0.215 moose km-2 (HDI 0.145, 0.286) in 2019 and 0.207 moose km-2 (HDI 0.144, 0.276) in 2020. Within-stratum estimates varied from year to year, likely an effect of the definition of suitable habitat and transect classification. Estimated ratios of bulls per cow and calves per cow fell within expected ranges for Shiras moose in Colorado but did vary across years. Thermal clutter created some impact on TIR video quality. Our study indicates that this approach can reliably estimate moose densities in forested and topographically complex environments, while maintaining greater aircraft altitude above ground level when compared with traditional aerial survey methods. This reduced disturbance to wildlife, impacts on wilderness, and improved aviator and aircraft passenger safety.
https://doi.org/10.5061/dryad.x69p8czt4
Description of the data and file structure
This repository contains spatial data, raw survey data, metadata, and analysis code supporting the manuscript: Abouelezz and Hobbs (2025). A high-altitude thermal infrared method for estimating moose abundance and demography in Rocky Mountain National Park, USA.
The dataset is organized into two primary components:
- Habitat Suitability Index (HSI) and transect spatial data
- Aerial transect survey data and associated analysis files
1. Habitat Suitability Index (HSI) Spatial Data
hsi.zip
Zipped raster file of the moose Habitat Suitability Index (HSI) projected across the study area, from which "less suitable" and "more suitable" transects were derived. The HSI was calculated using the equation: HSI = (SI1×SI2×SI3)1/3, where SI1, SI,2 and SI3 represent suitability indices derived from vegetation and landscape covariates described in the manuscript and Supplemental Information.
sample_grid.zip
Zipped ESRI shapefile containing the sample grid used to select aerial survey transects.
Spatial processing and geoprocessing workflows were conducted in ArcGIS Desktop 10.4 (Environmental Systems Research Institute, Redlands, CA, USA).
2. Aerial Transect Survey Data (Density, Population Size, and Demography)
2019_complete_data.csv
Raw line-distance double-observer survey data collected during fixed-wing high-resolution thermal infrared surveys conducted in July 2019. Includes transect identifiers, detection indicators, species identification, group composition (bulls, cows, calves), vegetation type, and associated survey attributes.
2020_complete_data.csv
Raw line-distance double-observer survey data collected in July 2020. Structure is consistent with the 2019 dataset.
2019_metadata.csv
Data dictionary describing variables contained in 2019_complete_data.csv.
2020_metadata.csv
Data dictionary describing variables contained in 2020_complete_data.csv.
ROMO_Moose_Elk_IR_Survey_Strata.csv
Lookup table linking sampled transects to habitat strata (“more suitable” and “less suitable”) used in stratified sampling and estimation procedures.
MooseAnalysis2020_revised.Rmd
Original R Markdown file used to generate density, population size, and demographic composition estimates from the raw survey data.
Data Notes
UTM Easting and Northing values in survey data have been redacted to protect sensitive wildlife locations. “R” indicates redacted spatial coordinates. “NA” indicates no coordinate was recorded during the survey.
Sharing/Access information
Original vegetation mapping information of Rocky Mountain National Park, Colorado, USA can be downloaded from: DataStore - Geospatial data for the Vegetation Mapping Inventory Project of Rocky Mountain National Park (nps.gov). Additional resources related to this dataset can be found here: Vegetation Inventory and Map for Rocky Mountain National Park (U.S. National Park Service) (nps.gov)
Code/Software
Spatial processing was completed using ArcGIS Desktop 10.4 (Environmental Systems Research Institute, Redlands, CA, USA)
Statistical analyses were conducted in R statistical software (R Core Team 2018) to produce density, population size, and demographic composition estimates.
