Data from: Assessing accuracy of GAP and LANDFIRE land cover datasets in winter habitats used by greater sage-grouse in Idaho and Wyoming, USA Dataset DOI: 10.7923/9CFT-9G84 Dataset published to support the following peer-reviewed manuscript: Title: Assessing accuracy of GAP and LANDFIRE land cover datasets in winter habitats used by greater sage-grouse in Idaho and Wyoming, USA Journal: Journal of Environmental Management Manuscript DOI: https://doi.org/10.1016/j.jenvman.2020.111720 Data use: License: CC-BY 4.0 - https://creativecommons.org/licenses/by/4.0/ Recommended Citation: Fremgen-Tarantino, M. R., Olsoy, P., Frye, G. G., Connelly, J. W., Krakauer, A. H., Patricelli, G. L., Child, A. W., & Sorensen Forbey, J. (2022). Data from: Assessing accuracy of GAP and LANDFIRE land cover datasets in winter habitats used by greater sage-grouse in Idaho and Wyoming, USA [Data set]. University of Idaho. https://doi.org/10.7923/9CFT-9G84 Data Files (confusion_matrices directory): confusion_matrices_2022.03.24.xlsx: Microsoft Excel table (.xlsx) with multiple tabs of data. Data was left in proprietary Microsoft format as to not compromise equations, linkages, and format. Individual tabs can be exported as txt or csv files, if needed for further analyses. xlsx tab structure: confusion_matrices_by_site: Confusion matrices detail the number of properly and improperly classified points within each study area, at each spatial scale (point/0m, 100 m, 1000 m and 5000 m). Matrices show the field classification (based on vegetation surveyed on the ground, e.g., ground-truthed data) and the classification in each data set. confusion_matrices_forage-rando: Confusion matrices detail the number of properly and improperly classified points within each study area, at each spatial scale (point/0m, 100 m, 1000 m and 5000 m). Matrices show the field classification (based on vegetation surveyed on the ground, e.g., ground-truthed data) and the classification in each data set. Matrices are divided into separate tables for forage sites (used) versus random points (available) to evaluate how classification errors may or may not differ in sites that are or are not winter forage locations caluculating_kappa_errors: This tab summarizes the data from the confusion matrices and uses the observed (Po) and expected (Pe) accuracy to calculate kappa. Site: Categorical assignments based on study area Values: Brown's Bench: Brown's Bench, ID Craters: Craters, ID Raft River: Raft River, ID Wyoming: Lander, WY Data set: Categorical assignment based on land cover classification analyzed Values: GAP LANDFIRE, community LANDFIRE, species Spatial Scale: Categorical assignment based on the spatial scale that land cover classification was analyzed for, based on the radius of a circle around the data point Values: 0 m 100 m 1 km (1000 m) 5 km (5000 m) Observed: Observed accuracy of the land cover classification as calculated from either "confusion_matrices_by_site" or confusion_matrices_forage-rando" tabs; continuous numerical Expected: Expected accuracy of the land cover classification as calculated from either "confusion_matrices_by_site" or "confusion_matrices_forage-rando" tabs; continuous numerical N: Number of points with associated vegetation data from the field transects, within a study area; integer Step 1 SD: Formula for calculating standard deviation (first step, see formula); continuous numerical Standard deviation: Standard deviation, calculated from "Step 1 SD"; previous column); continuous numerical Standard Error: Standard error of kappa value, calculated using standard formulas; continuous numerical SE for manuscript: Standard error of kappa raw value; continuous numerical Kappa: Kappa value, as calculated in the "confusion_matrices_by_site" or "confusion_matrices_forage-rando" tabs, using standard equations; continuous numerical bounded by 1 and -1 table_3: Summary data from "confusion_matrices_by_site" and "calculating_kappa_errors" to be published in manuscript. First set of columns (A:H) represent the actual data reported within the manuscript. Second set of columns (J:S) are calculation steps and equations to generate summary data. Study Area: Categorical assignment based on name of study area Values: Brown's Bench: Brown's Bench, ID Craters: Craters, ID Raft River: Raft River, ID Wyoming: Lander, WY Land Cover Dataset: Categorical assignment made based on land cover classification analyzed Values: GAP: GAP LANDFIRE: LANDFIRE Scale: Categorical assignment based on the scale used in assessment of land cover data Values: Ecological systems: GAP Community-level: LANDFIRE community-level Species-level: LANDFIRE species-level Spatial Scale: Categorical assignment based on the Spatial scale that land cover classification was analyzed, based on the radius of a circle around the data point Values: 0 m 100 m 1 km (1000 m) 5 km (5000 m) Overall Accuracy (%; SE): Overall accuracy and standard error (as a percent) as calculated with columns P:S. Raw Data is linked from "calculating_kappa_errors" tab, using standard equations. Kappa (SE): Calculated kappa value (continuous numerical number bounded by 1 and -1) and standard error as calculated using columns J:O. Raw data is linked from "calculating_kappa_errors" tab, using standard equations. Kappa Agreement Level: Categorical assignment based on kappa agreement between the field data and the land cover dataset Values: Strong: kappa > 0.8 Moderate: Kappa 0.4-0.8 Poor: kappa <0.4 Accuracy Rank (based on Overall Accuracy): Accuracy rank is how well the land cover dataset performed compared to other datasets, spatial scales, and study areas based on overall accuracy; integer (rank). [Column I]: NULL Overall accuracy raw: raw overall accuracy value from "calculating_kappa_errors" tab overall accuracy round (%): overall accuracy value as a percent rounded to one decimal place for reporting in manuscript. kappa raw: raw calculated kappa value from "calculating_kappa_errors" tab kappa round: rounded "kappa raw" value to two decimal places for reporting in manuscript kappa se raw: calculated standard error value for each kappa from "calculating_kappa_errors" tab kappa se round: "kappa SE raw" rounded to two decimal places for reporting in manuscript overall accuracy sample size: sample size of each study area for calculating overall accuracy SE overall accuracy raw: raw overall accuracy value from "calculating_kappa_errors" tab overall accuracy SE raw: standard error of overall accuracy calculated using standard equations overall accuracy SE round (percent): standard error of overall accuracy as a percent rounded to one decimal place for reporting in manuscript. table_4: Summary data from "confusion_matrices_forage-rando" and "calculating_kappa_errors" to be published in manuscript. First set of columns (A:J) represent the actual data reported within the manuscript. Second set of columns (L:AA) are calculation steps and equations to generate summary data. Data Set: Categorical land cover classification analyzed Values: GAP LANDFIRE Scale: Categorical assignment based on the scale used in assessment of land cover data Values: Ecological systems: GAP Community-level: LANDFIRE community-level Species-level: LANDFIRE species-level Study Area: Categorical assignment based on name of study area Values: Brown's Bench: Brown's Bench, ID Craters: Craters, ID Raft River: Raft River, ID Wyoming: Lander, WY Forage Patches: Overall Accuracy (%: SE): Overall accuracy and standard error (as a percent) as calculated with columns R:V. Raw Data is linked from "calculating_kappa_errors" tab, using standard equations. Kappa (SE): Calculated kappa value (continuous numerical number bounded by 1 and -1) and standard error as calculated using columns L:N. Raw data is linked from "calculating_kappa_errors" tab, using standard equations. Random Patches: Overall Accuracy (%: SE): Overall accuracy and standard error (as a percent) as calculated with columns W:AA. Raw Data is linked from "calculating_kappa_errors" tab, using standard equations. Kappa (SE): Calculated kappa value (continuous numerical number bounded by 1 and -1) and standard error as calculated using columns O:Q. Raw data is linked from "calculating_kappa_errors" tab, using standard equations. Higher Overall Accuracy: Categorical assignment based on comparison of overall accuracy of used (forage) and available (random) points by listing the highest overall accuracy Values: Forage Random Equal Higher Kappa: Categorical assignment based on comparison of the raw kappa values (columns L and O) of used (forage) and available (random) points by listing the highest overall accuracy Values: Forage Random Equal graphs_fig5: Data used to create graphs showing the level of overall accuracy for each land cover classification based on the spatial scale of each classification, by study area (for the four largest study areas). Site: Categorical assignment based on study area name Values: Brown's Bench: Brown's Bench, ID Craters: Craters, ID Raft River: Raft River, ID Wyoming: Lander, WY Data set: Categorical assignment based on land cover classification analyzed Values: GAP LANDFIRE, community LANDFIRE, species Spatial Scale: Integer assignment based on the spatial scale that land cover classification was analyzed for, based on the radius of a circle around the data point Values: 0: 0 m 100: 100 m 1000: 1000 m/1 km 5000: 5000 m/5 km Overall accuracy: Overall accuracy as a percent, as calculated in "confusion_matrices_by_site" tab, using standard equations; continuous numerical correlation_summaries: Summary of statistical tests (calculated in R using "AccuracyByRadius.R" script) evaluating if data are correlated at different spatial scales. Site: Categorical assignment based on name of study area Values: Brown's Bench: Brown's Bench, ID Craters: Craters, ID Raft River: Raft River, ID Wyoming: Lander, WY Data set: Categorical assignment based on land cover classification analyzed Values: GAP: GAP LF Community: LANDFIRE, community-level LF Species: LANDFIRE, species-level Distribution: Categorical assignment based on the results of Shapiro-Wilk normality test Values: Normal: normally distributed (Shapiro-Wilk p>0.05) Not normal: not normally distributed (Shapiro-Wilk p<0.05) Test: Categorical assignment of correlation test used to analyze data based on distribution of data Values: Pearson: Normal distribution Spearman rank: Not normal distribution S or T: Statistical value as calculated in R (S for Spearman rank and t for Pearson test); continuous numerical p-value: p-value for the test as calculated in R; continuous numerical df: Degrees of freedom for test; integer correlated?: Boolean response for results of respective Pearson or Spearman Rank correlation test in R. Values: Yes: correlation was significant; p-value < 0.05 No: correlation was not significant; p-value > 0.05