Lidar derived Basal Area Weighted Height, Minnesota
Go to Section:- Overview
- Data Quality
- Data Organization
- Coordinate System
- Attributes
- Distribution - Get Data
- Metadata Reference
Section 1: Overview
Originator:Minnesota Department of Natural Resources (DNR)
Title:
Lidar derived Basal Area Weighted Height, Minnesota
Abstract:
The Basal Area Weighted Height (BAWHT) raster layer (at 20-m spatial resolution) is a modelled forest inventory attribute based on integration of Minnesota's 3DEP USGS Lidar and temporally coincident plot based inventory (PBI) data collected by the MN DNR Division of Forestry Resource Assessment office.
For more information about all metrics and models, see the Resource Assessment Glossary and Frequently Asked Questions documents, found on the Resource Assessment Program website.
Frequently Asked Questions
Resource Assessment Glossary
Resource Assessment | MN DNR
Purpose:
BAWHT is a spatially explicit raster output for the entire Lidar Acquisition Block (LAB). Each pixel in the raster represents height of the average tree in a stand. The spatial inventory modeling of BAWHT, also called Lorey's height, required field inventory data and co-located lidar data from many 0.1 acre PBI plots over the LAB. The plot-level BAWHT is obtained by multiplying the total height and basal area of individual sample trees, and then dividing the sum of the products by the total plot-level basal area. Basal area weighted heights are calculated for each of the sample plots and the lidar derived metrics for the plots are used as predictors to build a relationship (regression model) that is finally extended over the entire area of interest using the most significant lidar predictors at 20 m spatial resolution.
Time Period of Content Date:
Currentness Reference:
The 3DEP USGS Lidar data used to produce this data was acquired in Lidar Acquisition Blocks at various time periods in the state, starting in 2021, and up to 2025.
Progress:
In work
Maintenance and Update Frequency:
None Planned
Spatial Extent of Data:
Statewide
Bounding Coordinates:
-97.23
-89.53
49.37
43.5
Place Keywords:
Minnesota
Theme Keywords:
lidar, remote sensing, forestry, 3DEP USGS
Theme Keyword Thesaurus:
Access Constraints:
None. Please see 'Distribution Info' for details.
Use Constraints:
None. Users are advised to read the dataset's metadata thoroughly and to have background knowledge in forestry to understand appropriate use and data limitations.
Browse Graphic:
None available
Associated Data Sets:
Minnesota's 3DEP USGS Lidar, LAB extent (polygon feature class)
Section 2: Data Quality
Attribute Accuracy:
Logical Consistency:
Completeness:
The data included is for the Lidar Acquision Blocks where the data is available.
Horizontal Positional Accuracy:
N/A
Vertical Positional Accuracy:
N/A
Lineage:
The Lidar inputs are from certified 3DEP USGS data as referenced by
RLB (2021) Project Report
https://rockyweb.usgs.gov/vdelivery/Datasets/Staged/Elevation/metadata/MN_RainyLake_2020_B20/USGS_MN_RainyLake_2020_B20_Project_Report.pdf
LSB (2021) Project Report
https://rockyweb.usgs.gov/vdelivery/Datasets/Staged/Elevation/metadata/MN_LakeSuperior_2021_B21/USGS_MN_LakeSuperior_2021_B21_Project_Report.pdf
PiCo (2019) Project Report
https://rockyweb.usgs.gov/vdelivery/Datasets/Staged/Elevation/metadata/WI_Oshkosh_3Rivers_2018_D19/WI_Oshkosh_3Rivers_2018_D19_WP_Report.pdf
The mask created for Lidar Derived forest inventory models is based on two parameters. First, the areas in the forest inventory models that have negative values. The negative values in a forest inventory model correspond to areas where lidar returns are absent, or are in a condition that was not sampled during field collection. Second, the mask omits areas where less than 10% of the first returns are above 3m from the ground.
BAWHT is modeled using multivariate relationships between field measured plot (0.1 acre) data and lidar-derived grid metrics (at 20 m resolution). A series of canopy grid metrics, or descriptive statistics, are created from the lidar data in raster format. Plot Based Inventory (PBI) plot data trains the model in Random Forest to determine the metrics that contribute the most to prediction of the forest inventory metrics. A multiple linear regression model is developed based on the most valuable grid metrics. The predictor variables selected in this way are further subjected to stepwise backward and forward regression to have a parsimonious set of significant predictors that are finally used to fit a weighted multiple linear regression model (also called generalized least square model, ‘gls’) in the ‘nlme’ package in R statistical programming. In general, the model weight is based on a predictor that yields highest correlation with the response inventory variable. The weighted multiple linear regression model is then extended spatially (wall-to-wall) using the same parsimonious set of predictors as grid metrics at 20 m resolution.
The version of the BAWHT Model refers to the sequential order of running the Model on updated PBI data.
The multiple linear regression model for BAWHT in the following LABs is in the form:
where a0 is intercept and a1-a5 are linear model coefficients
RLB (2021) - Model v2
BAWHT = a0 + (a1 * ElevMax.tif) + (a2 * ElevAv.tif) + (a4 * ElevSD.tif) - (a3 * strata3.tif)
LSB (2021) - Model v2
BAWHT = a0 + (a1 * ElevMax.tif) + (a2 * strata3.tif) + (a3 * cov3m1st.tif) + (a4 * strata4.tif) + (a5 * ElevSD.tif) + (a6 * ElevP25.tif)
PiCo (2019) - Model v3
BAWHT = a0 + (a1 * ElevMax.tif) + (a2 * strata4.tif) + (a3 * strata5.tif) + (a4 * ElevVar.tif) + (a5 *strata6.tif)
Section 3: Spatial Data Organization (not used in this metadata)
Section 4: Coordinate System
Horizontal Coordinate Scheme:
Universal Transverse Mercator
UTM Zone Number:
15
Horizontal Datum:
NAD83(2011)
Horizontal Units:
meters
Vertical Datum:
Vertical Units:
Depth Datum:
Depth Units:
Cell Width:
20m
Cell Height:
20m
Section 5: Attributes
Overview:
Value range 0+ represent the BAWHT in feet.
Detailed Citation:
Support for this project is provided by the State of Minnesota and the Minnesota Environment and Natural Resources Trust Fund (ENTRF) as recommended by the Legislative-Citizen Commission on Minnesota Resources (LCCMR).
Resource Assessment | MN DNR
Table Detail:
Section 6: Distribution
Publisher:
Minnesota Department of Natural Resources (DNR)
Publication Date:
01/10/2025
Contact Person Information:
Jason Langenecker,
Assistant Resource Assessment Supervisor for Field Services
Minnesota DNR - Forestry, Resource Assessment
483 Peterson rd
Grand Rapids,
MN
55744
Phone: 218-322-2519
Email: jason.langenecker@state.mn.us
Distributor's Data Set Identifier:
Ordering Instructions:
Lidar data may be accessed directly from the 3DEP USGS program via the National Map https://apps.nationalmap.gov/viewer/.
For access to the Lidar derived forest inventory raster datasets, send a message to the Distribution Contact. Online Linkage:
None available
Section 7: Metadata Reference
Metadata Date:
01/10/2025
Metadata Standard Name:
Minnesota Geographic Metadata Guidelines
Metadata Standard Version:
1.2
Metadata Standard Online Linkage: