University of Florida (UF) and Florida International University (FIU)
researchers have performed a number of projects during the past two years to
test and demonstrate the capabilities of ALSM techniques, to improve the data
reduction and analysis procedures and to explore the most useful way to provide
the results to users. The projects have included mapping of sandy beaches
subject to erosion and tropical storm damage and environmentally sensitive
marshlands. Early results of these projects indicate that the system can yield
heights accurate to 5 to 10 centimeters, and detect linear topographic features
with height differences of 2 to 5 centimeters. The UF and the International
Hurricane Center, Florida International University, purchased an ALSM
system in 1999, and
researchers interested in using ALSM in their research are invited to contact
the authors.
INTRODUCTION
The basic concepts of airborne laser swath mapping are simple. A pulsed laser
ranging system is mounted in an aircraft equipped with a precise kinematic
Global Positioning System (GPS) receiver and an inertial navigation system.
Solid state lasers are now available that can produce
thousands of pulses per second, each pulse having a duration of a few
nanoseconds (10-9 seconds). Light travels approximately 30
centimeters in one nanosecond. By accurately timing the round trip travel time
of the light pulses from the aircraft to the ground (water, foliage, buildings
or other surface features) it is possible to determine the range with a
precision of one centimeter or better. Using a rotating mirror inside the laser
transmitter, the laser pulses can be made to sweep through an angle, tracing out
a line on the ground. By reversing the direction of rotation at a selected
angular interval, the laser pulses can be made to scan back and forth along a
line. When such a laser ranging system is mounted in a aircraft with the scan
line perpendicular to the direction of flight, it produces a saw tooth pattern
of ranges within a strip centered directly along the flight path (Figure 1). The
width of the strip or "swath" covered by the ranges, and the spacing
between measurement points, depends on the scan angle of the laser ranging
system and the airplane height. Using a light twin or single engine aircraft,
typical operating parameters are flying speeds of 200 to 250 kilometers per hour
(55 to 70 meters per second), flying heights of 300 to 1000 meters, scan angles
up to ±20 degrees, and pulse rates of  2000 to
5000 pulses per second. These parameters can be selected to yield a measurement
point every few meters, with a footprint of 10 to 15 centimeters, providing
enough information to create a digital terrain model (DTM) adequate for most
applications, including the mapping of storm damage to beaches, in a single
pass.
After a flight the precise position of the aircraft at the exact epoch of
each range measurement is computed relative to nearby GPS ground stations using
phase differenced kinematic Global Positioning System (DGPS) techniques. The
laser ranging vectors are added to the aircraft positions to derive three
dimensional X,Y,Z coordinates of each ground point. As long as the airplane is
within 10 to 20 kilometers of a ground base station it is possible to determine
the position of the airplane within about 3 to 5 centimeters using single
frequency GPS observations and the broadcast ephemeris. However, as the distance
from the aircraft to the nearest ground GPS station increases the error in the
aircraft coordinates will increase, and it becomes important to use a precise
ephemeris and dual frequency observations to remove errors caused by the
ionosphere. The X,Y,Z coordinates of the ground points can be processed using a
number of commercially available software packages to produce a Digital Terrain
Model (DTM) and many other products such as shaded relief maps, contour maps,
cross sections, surface profiles, and to compute engineering quantities such as
land areas and changes in volumes associated with civil engineering works or
storm erosion.
RECENT PROJECTS
Project LASER
In early October 1995 hurricane Opal struck the Gulf Coast along the
panhandle of Florida, doing extensive damage to upland structures and to the
beaches and dunes. Researchers at the University of Florida organized a project
to map the area hit by hurricane Opal using airborne laser swath mapping
techniques1,2,3,4. This multi-agency effort was named project LASER
and the field data collection portion of project LASER was conducted during the
period October 15, 16 and 17, 1996. A light twin engine aircraft, operated by
the Florida Department of Transportation (FDOT) for airborne photogrammetry and
already equipped with a dual frequency GPS antenna, camera port, and mountings
that could easily accommodate the sensor package and electronics rack of the
Optech Inc. ALTM 1020 laser ranging system (Figure 2), was used for the data
collection. Table 1. summarizes the specifications of the ALTM 1020 system. The
entire stretch of beach from Mexico Beach, Florida, to the west tip of Perdido
Key, Alabama, more than 250 kilometers in length, was mapped in both directions
in just over two hours. The nominal operating parameters were: flying height of
350 meters, scan rate of 25 cycles per second, scan angle of plus and minus 15
degrees, laser pulse rate of 5000 pps.
The raw GPS, inertial navigation, and laser ranging data collected during the
flight were stored on 8 mm cassette tapes. After completion of each flight the
data from the GPS reference ground stations were brought together with the data
from the aircraft, and Optech personnel used proprietary computer programs to
combine the GPS, inertial navigation, and laser ranging data to compute the
coordinates of the surface points. The results were ASCII files containing lists
of UTM Eastings, Northings, ellipsoid heights, and times for all the successful
range measurements collected during the observing window. The percentage of
successful range measurements, as a function to the number of laser pulses
transmitted was typically more than 90 percent over the high reflectivity beach
areas, but dropped quickly over water, reaching fewer than 10 percent when the
point on the water surface was more than 6 to 8 degrees off nadir.
Figure 3 shows an area of beach and near shore water at the western end of
Perdido Key, Alabama, with the locations of the surface points from which
reflections of sufficient strength were received to obtain range measurements.
The coverage is very dense over the relatively diffusely reflective white sand
beaches and dunes. Over the near shore water many of the returns were
insufficient to measure the range because the reflection is more directional
from the surface of the water, and often does not reflect in the direction of
the aircraft. Figure 4 is a shaded relief map of the same area shown in Figure
3, with a number of features labeled. Figure 5 is the same area as Figure 4, but
the results are presented as a false color contour map. In both Figures 4 and 5
waves propagating past the tip of the land (past a jetty) are quite apparent.
Figure 6 shows a cross section taken perpendicular to the wave fronts showing
the heights and spatial separation of the waves.
Table 1. ALTM 1020 Specifications
| Airborne Module |
Operating altitude: |
330 - 1000 m nominal |
| Range accuracy: |
15 cm single-shot |
| Range resolution: |
1 cm |
| Scan angle: |
Variable from 0 to ±20 |
| Swath width: |
Variable from 0 to 0.68 x altitude |
| Angle accuracy: |
0.05 |
| Angle resolution: |
0.01 |
| Scan frequency: |
Variable; depends on scan angle; e.g., 30 Hz for ±20 scan, 50 Hz for
±10 scan |
| Roll and Pitch Accuracy: |
0.04 |
| Heading Accuracy: |
0.05 |
| Supported GPS Receivers: |
Sercel NR103T, Trimble 4000SSE, or Astech Z12 |
| Laser wavelength: |
1047 nm |
| Laser repetition rate: |
100 Hz to 5kHz |
| Beam divergence: |
0.25 mrad |
| Laser classification |
Class IV laser product (FDA CFR 21) |
| Eyesafe range: |
308 m (single shot) |
| Power requirements: |
28 VDC @ 15 A |
| Operating temp.: |
10 - 35C |
| Humidity: |
0 - 95% non-condensing |
| Sensor: |
Fits all existing camera mounts, or can be directly mounted to the
floor |
| Dimensions: |
290 x 250 x 430 mm |
| Weight: |
11.4 kg (25 lbs) |
| Control rack: |
1 stackable vibration-isolated transportable case |
| Dimensions: |
60 x 60 x 65 cm, excluding GPS |
| Weight: |
45.4 kg (100 lbs) including shipping covers and cables |
| Video Output: |
NTSC or PAL (annotated video out) |
| Data Storage: |
8 mm digital data tape |
| Ground based module |
Software: |
ALTM 1020 GBPP |
| GPS receiver: |
Sercel NR103T, Trimble 4000SSE, or Astech Z12 |
| Personal computer:
(min. requirements)
|
Pentium laptop, 1 Gb hard disk
8 Mbyte RAM, SVGA graphics
Exabyte 8505 tape drive, SCSI adapter
|
Waccasassa Marsh
During November 1997, airborne laser swath mapping was used to map portions
of Waccasassa Marsh, near Cedar Key, Florida. Previous projects had demonstrated
that the technology worked well in bare beach areas and sparsely vegetated
dunes, but it was questionable if the laser pulses would penetrate the dense
vegetation in the marsh.
The Waccasassa flights were made with a Cessna 206 single engine aircraft.
The aircraft was equipped with a real time DGPS receiver and navigation aid that
enabled the pilot to fly preplanned patterns to within a few tens of meters. The
nominal flying height was 375 meters with an air speed of approximately 50
meters per second. The laser ranging system operated at 3000 pulses per second
and the scan pattern was a sawtooth with 25 cycles per second, with a swath
width of 180 meters. The raw mapping data were processed by Optech Inc.
personnel using proprietary software to produce north, east and up coordinates
relative to the WGS 84 UTM coordinate reference frame. The data delivered
included files with all ranges and files which were processed to
"remove" the vegetation. UF personnel processed the data using both
in-house software GATORPLEX and commercial software to produce a variety of
products, including three dimensional shaded relief maps, contour maps, and
selected profiles. An example of the initial results is shown in Figure 7.
Pinellas County
The UF, FIU and Pinellas County have a 3 year cooperative research program to
explore the application of ALSM to County programs and projects, including a
project to check the Federal Emergency Management Agency (FEMA) flood zone maps
for the County. During the first phase of this program ALSM data were collected
using an Optech Inc. Model ALTM 1020 system, mounted in a Cessna 206 single
engine aircraft equipped with a Starlink real time GPS receiver and Light Bar to
assist the pilots in flying straight parallel flight lines. Several different
flight line layouts were considered but the unusual shape and complex patterns
of land and water areas throughout the county made attempts to separate high
priority (flood zones) areas from the general coverage difficult, and it was
decided to use a simple pattern. The pattern selected was 77 parallel
north-south flight lines of varying lengths providing total coverage of land
within the County. The combined length of the 77 flight lines totals
approximately 3700 kilometers. At a nominal ground speed of 60 meters per second
it required more than 20 hours of on-line data collection. At 5000 laser pulses
per second, more than 350 million measurements were made, just for the basic
coverage, exclusive of special observations for calibration and validation of
the ALSM results.
The general coverage flights were all planned for a nominal altitude of 600
meters, an air speed of 60 meters per second (135 miles per hour), a scan angle
of plus and minus 20 degrees, a scan frequency of 12 cycles per second, and a
laser pulse rate of 5000 pulses per second. The nominal swath width, i.e., the
ground coverage from each swath, was 450 meters. The nominal overlap between
adjacent swaths was 80 meters. The actual flight lines vary significantly from
those planned largely because of wind effects on the aircraft and other lesser
errors including errors in the real time GPS navigation of the aircraft. The
altitude varies tens of meters because of atmospheric affects on the aircraft
and pilot skill even in the best of conditions, and in a few instances the
flight lines were intentionally reduced in order to continue operations in areas
with clouds at or very near the 600 meter level.
Two additional sets of flight lines were planned; 1) a pair of lines
following immediately along the beach line, and 2) eight transverse flight lines
spaced at several mile intervals covering the north-south extent of the County.
The purposes of the beach line flights were to obtain continuous coverage along
the beach to improve the delineation of the beach up to and a few hundred meter
beyond the line of the sea wall, and to provide an optimum data set for
comparison with the ground survey beach cross sections to calibrate the laser
swath mapping observational data. To achieve the best possible detection of the
sea wall the scan angle was reduced to plus and minus 15 degrees with a scan
rate of 4 cycles per second and the nominal flying height was reduced to 450
meters. The purpose of the transverse flights was to check the primary
north-south trajectories for systematic errors such as day to day offsets or
slopes. The laser system parameters were selected to maximize foliage
penetration, using a scan angle of plus and minus 15 degrees and a scan rate of
4 cycles per second. Figure 8 shows some early results from the Pinellas County
project.
ACCURACY OF LASER SWATH MAPPING
The position of the aircraft obtained by differential kinematic GPS is
determined in an earth centered earth fixed (ECEF) X,Y,Z cartesian coordinate
system. The origin, scale and orientation of the reference frame depends on the
satellite ephemerides and the coordinates of the GPS ground reference stations,
which should be consistent with one another. Adding the laser ranging vectors to
the aircraft positions yields ECEF X,Y,Z cartesian coordinates for each of the
ground points surveyed. However, most users need (or prefer) to work in a
reference system such as a national geodetic reference system or a state plane
coordinate system, where the coordinates more nearly represent local north,
east, and up. Also, most users prefer to use orthometric heights, i.e., heights
above the geoid, because they often want to mix or compare the heights with
those obtained from conventional (spirit) leveling. To transform the cartesian
coordinates obtained from airborne laser swath mapping to a system involving
orthometric heights, the undulation of the geoid must be known at each point. In
the United States geoid models are produced by the National Geodetic Survey(NGS)
, and the models are periodically improved as additional data become available5.
The ALSM data used in this paper were converted to orthometric heights using NGS
geoid model GEOID96.
Geoid models often are the least accurate immediately along shore lines,
because the geoid changes rapidly in coastal
areas and because the offshore gravity data required to develop the geoid model,
usually collected by aircraft or ships, is often sparse and of lower accuracy
than elsewhere. Errors in the geoid model could cause errors in the orthometric
heights of a decimeter or more in the Florida panhandle area. For applications
involving only changes in the surface, such as beach erosion, this is not a
problem for repeat surveys using only airborne laser swath mapping data, because
the geoid can be assumed to remain fixed, and the changes can be accurately
determined. To check for systematic errors and estimate the accuracy of the
laser mapping, the results were compared to cross sections of the beach obtained
by classical spirit leveling (Figure 9). Comparison of the several laser and
survey profiles spread widely spread along the beach show the laser results to
be accurate to the 5 to 10 cm level.
Table 2. Specifications of the ALTM 1210 Currently Under
development by Optech, Inc.
 |
Operating altitude: |
330 - 2000 m |
| Range accuracy: |
10 cm single-shot |
| Range resolution: |
1 cm |
| Relative accuracy |
2-4 cm @2khz, 5-10 cm @10 khz |
| Scan angle: |
Variable from 0 to ±20 |
| Options |
Intensity data |
| Simultaneous first and last pulse measurements |
| Extended altitude (up to 2000m) operation |
| Swath width: |
Variable from 0 to 0.68 x altitude |
| Angle accuracy: |
0.05 |
| Angle resolution: |
0.01 |
| Scan frequency: |
Variable; depends on scan angle; e.g., 30 Hz for ±20 scan, 50 Hz for
±10 scan |
| Roll and Pitch Accuracy: |
0.04 |
| Heading Accuracy: |
0.05 |
| Supported GPS Receivers: |
Astech Z12 or Trimble 4000SSE |
| Laser wavelength: |
1047 nm |
| Laser repetition rate: |
100 Hz to 10kHz |
| Beam divergence: |
0.30 mrad |
| Laser classification |
Class IV laser product (FDA CFR 21) |
| Eyesafe range: |
308 m (single shot) |
| Power requirements: |
28 VDC @ 30 A |
| Operating temp.: |
10 - 35C |
| Humidity: |
0 - 95% non-condensing |
| Sensor: |
Fits all existing camera mounts, or can be directly mounted to the
floor in a small single engine aircraft such as Cessna 172 |
| Dimensions: |
290 x 250 x 500 mm |
| Weight: |
11.4 kg (25 lbs) |
| Control rack: |
1 stackable vibration-isolated transportable case |
| Dimensions: |
60 x 60 x 75 cm, excluding GPS |
| Weight: |
50 kg including shipping covers and cables |
| Video Output: |
NTSC or PAL (annotated video out) |
| Data Storage: |
12 hour capacity (8 mm digital data tape) |
 |
Software: |
ALTM 1020 GBPP |
| GPS receiver: |
Astech Z12 or Trimble 4000SSE |
| Personal computer:
(min. requirements)
|
Pentium laptop, 1 GB hard disk, 8 Mbyte RAM
SVGA graphics, Exabyte 8505 tape drive, SCSI adapter
|
FUTURE OF THE ALSM TECHNOLOGY
The UF and the International Hurricane Center, Florida International
University, are purchasing an ALSM system which is expected to be operational in
the first quarter of 1999. Specifications of this new instrumentation are shown
in Table 2. UF and FIU have also purchased a Cessna 337 (in-line twin engine)
which is being modified to house the ALTM 1012G system. The Universities plan to
make this system available to collect data for research at other organizations
and institutions on a cost reimbursable basis. Researchers interested in using
ALSM are invited to contact the authors.
CONCLUSIONS
ALSM technology has been used to accurately map hundreds of kilometers of
beaches and near shore water surfaces in a few hours. This new capability opens
for the first time the possibility to regularly monitor all the beaches of the
nation, to map specific beaches immediately prior to and after major storms to
obtain accurate quantitative information about the extent of the damage, to
follow the natural recovery of beaches after storms and to monitor the effects
of actions such as beach nourishment or other anthropogenic modifications. Laser
mapping of near shore waters can be used to determine the direction wave spectra
of a coastal area to determine the characteristics of storms that result in the
most damage to specific areas. There is no doubt that airborne laser swath
mapping technology will continue to improve as higher pulse rate lasers become
available, kinematic GPS techniques are refined, and more powerful computers and
software are developed. But, the technology has already reached a point where
the time and costs are so small relative to the benefits to be derived that it
should be placed into operational use immediately.
REFERENCES
1. Carter, W.E. and R.L. Shrestha, "Airborne Laser Swath Mapping:
Instant Snapshots of Our Changing Beaches," In Proceedings of the
Fourth International Conference: Remote Sensing for Marine and Coastal
Environments," Environmental Research Institute of Michigan (ERIM),
P.O. Box 134001, Ann Arbor, MI 48113-4001, USA, Vol. I, pp. 298 - 307, 1997.
2. Carter, W. E, R. L. Shrestha, P.Y. Thompson, and R. G. Dean; "Project
LASER: Final Report to FDEP," Department of Civil Engineering,
University of Florida, Gainesville, FL 32611, pp 27, April 4, 1997.
3. Carter, W. E, R. L. Shrestha, and P.Y. Thompson; "Project LASER:
Final Report to Florida Department of Transportation," Department of
Civil Engineering, University of Florida, Gainesville, FL 32611, pp 31,
June 16, 1997.
4 Shrestha, R.L., W.E. Carter, P.Y. Thompson, R.G. Dean and H. Harrell,
"Coastal & Highway Mapping by Airborne Laser Swath Mapping
Technology," presented and published in The Proceedings of Third
International Airborne Remote Sensing Conference and Exhibition,
Copenhagen, Denmark, Vol. I, pp. 632 - 639, 1997
5. Milbert, D.S., Computing GPS-Derived Orthometric Heights with Geoid90
Height Model, Presented at ACSM Fall Meeting, GIS/LIS, Atlanta, GA, 1991.
Copyright © 2000 University of
Florida holds all rights reserved. Reproduction in whole or
in part in any form or medium without the express written
permission of The University of Florida is prohibited. (Article reprinted with permission of Bill Carter).
|
