LiDAR Quality Levels

A LiDAR point cloud often is described by its Quality Level (QL), which is comprised of the nominal pulse spacing / density (NPS / NPD) and the relative and absolute vertical accuracy of the point cloud and derived digital elevation models. NPS and NPD refer to the point density of single swath, single instrument, first-return-only data. The aggregate nominal pulse spacing / density (ANPS / ANPD) describes the net overall pulse spacing and density including overlapping swaths and multiple passes of the LiDAR system.

Table 1 and Table 2 (from the USGS LiDAR Base Specification v1.3) describe LiDAR Quality Levels with regard to ANPS / ANPD, relative accuracy, and absolute accuracy.

Table 1
Aggregate Nominal Pulse Spacing and Density

Quality Level
Aggregate Nominal Pulse
Spacing (m)
Aggregate Nominal Pulse
Density (pls/m2)
QL0
≤ 0.35
≥ 8.0
QL1
≤ 0.35
≥ 8.0
QL2
≤ 0.71
≥ 2.0
QL3
≤ 0.35
≥ 0.5

QL quality level  pls/m2 pulses per square meter  m meter   less than or equal to  greater than or equal to

Table 2
Relative Vertical Accuracy for LiDAR Swath Data

Quality Level
Smooth Surface
Repeatability, RMSD, (m)
Swath Overlap
Difference, RMSD, (m)
QL0
≤ 0.03
≤ 0.04
QL1
≤ 0.03
≤ 0.08
QL2
≤ 0.08
≤ 0.08
QL3
≤ 0.12
≤ 0.16

QL quality level  pls/m2 pulses per square meter  m meter  less than or equal to  greater than or equal to

Table 3 and Table 4 describe land cover classes used for reporting absolute accuracy of the LiDAR point cloud.

Table 3
Land Cover Classes

Class
Land Cover Class or Description
Previous Reporting Group
Current Reporting Group
1
Clear or open, bare earth, low grass; for example: sand, rock, dirt, plowed fields, lawns, golf courses
FVA
NVA
2
Urban areas; for example: tall, dense man-made structures
SVA
NVA
3
Tall grass, tall weeds and crops: for example, hay, corn, wheat fields
SVA
VVA
4
Brush lands and short trees; for example: chaparrals, mesquite
SVA
VVA
5
Forested areas, fully covered by trees; for example: hardwoods, conifers, mixed forests
SVA
VVA
6
Sawgrass
n/a
n/a
7
Mangrove and swamps
n/a
n/a

FVA Fundamental Vertical Accuracy SVA Supplemental Vertical accuracy NVA Nonvegetated Vertical Accuracy VVA vegetated vertical accuracy n/a not
applicable or not available

Table 4
Absolute Vertical Accuracy for LiDAR Data and Digital Elevation Models

Quality Level
RMSE (Non Vegetated) (m)
NVA at the 95% Confidence Level (m)
VVA at the 95th Percentile (m)
QL0
≤ 0.050
≤ 0.098
≤ 0.15
QL1
≤ 0.100
≤ 0.196
≤ 0.30
QL2
≤ 0.100
≤ 0.196
≤ 0.30
QL3
≤ 0.200
≤ 0.392
≤ 0.60

QL quality level RMSE root mean square error in the z direction NVA non-vegetated vertical accuracy VVA vegetated vertical accuracy m meter less than or equal to

Table 5 Describes the minimum classification of the LiDAR point cloud for the USGS 3D Elevation Program. Table 6 Describes the requirements for the minimum cell size for the digital elevation models that are derived from the bare-earth classified point cloud.

Table 5
Minimum LiDAR Data Classification Scheme

Code
Description
1
Processed, but unclassified
2
Bare earth
7
Low noise
9
Water
17
Bridge deck
18
High noise
20
Ignored Ground (typically breakline proximity)
21
Snow (if present and identifiable)
22
Temporal Exclusion (typically non-favored data in intertidal zones)

Table 6
Minimum Digital Elevation Model Cell Size

Quality Level
Minimum Cell Size (m)
Minimum Cell Size (ft)
QL0
0.5
1.0
QL1
0.5
1.0
QL2
1
2
QL3
2
5

QL quality level   m meter   ft feet