One year's experience with a recreation-grade GPS receiver

1
One year's experience with a recreation-grade GPS
receiver
Pete Bettinger Warnell School of Forestry and
Natural Resources University of Georgia Athens,
GA Songlin Fei Department of Forestry University
of Kentucky Lexington, KY
7th Southern Forestry and Natural Resources GIS
Conference
2
Overview There are three general classes of GPS
receivers used in natural resource
management Consumer- or recreation-grade
Mapping-grade Survey grade
Recreation-grade receivers generally provide the
least accurate positional information - between
3 m and 10 m accuracy under optimal
conditions. However, recreation-grade receivers
have become popular among many outdoors
enthusiasts, and this popularity has likely
influenced the wide variety of inexpensive GPS
receivers available on the market today.
7th Southern Forestry and Natural Resources GIS
Conference
3
Overview Which type of receiver to choose? It
depends on your application, and your cost
considerations. If single-fix errors of 5-20 m
are acceptable, then a recreation-grade receiver
would suffice. Also, recreation-grade may be
acceptable for locating field points when you
know you will be in one location for some time
(allowing the capture of multiple fixes). When
mapping large areas, where the error concerning
the perimeter is small relative to the size of
the area, then a recreation-grade receiver may
also be acceptable. Ultimately, the cost of
data collection and the desired accuracy of
referenced positions should be balanced.
7th Southern Forestry and Natural Resources GIS
Conference
4
Objectives The objectives of this research were
to compare the accuracy of horizontal position
locations in different timber types, and across
different seasons and months of the year. In
addition, the research was aimed at examining the
influence of air temperature, relative humidity,
atmospheric pressure, and planned positional
dilution of precision on the accuracy of
captured position locations.
7th Southern Forestry and Natural Resources GIS
Conference
5
Methods Three GPS test points were selected from
the 40 available at the Whitehall Forest GPS
Test Course in Athens, GA. The test points are
located in a young pine plantation (15 years
old), an older, natural pine stand (over 60
years old on average), and a mature hardwood
stand (over 60 years on average). The three
test points were visited once per day over the
course of a year, and 50 position fixes
(waypoints) were collected during each visit,
with each position fix captured within 2-3
seconds of the previous.
Young pine stand 15 years old 130 ft2 per acre
basal area 450 trees per acre Hardwood
stand 60-70 years old 88 ft2 per acre basal
area 144 trees per acre Older pine stand 60-70
years old 86 ft2 per acre basal area 59 trees
per acre
7th Southern Forestry and Natural Resources GIS
Conference
6
Methods
7th Southern Forestry and Natural Resources GIS
Conference
7
Methods A Garmin Oregon 300 was selected for use
in this study. WAAS was disabled for this study
to ensure that the data were collected under
consistent circumstances. Planned PDOP was
collected from Trimble GPS planning
software. Data were downloaded to a computer
using Minnesota DNR Garmin software. Some
warm-up time was necessary to ensure that a
sufficient number of satellites were available to
provide a reasonable position fix. Visits were
randomly selected, and each test point was
visited within about 2-3 minutes of the previous
point. During each visit, the GPS receiver was
plumbed over each control point, and the
operator stood on the north side of each as data
was collected.
7th Southern Forestry and Natural Resources GIS
Conference
8
Methods Accuracy of the horizontal positions
collected was evaluated by using the root mean
square error (RMSE). RMSE is the raw difference
between collected measurements and the control
points, and places greater weight on larger
errors since the error term is squared. A
correlation analysis was performed between the
RMSE values and the following Forest
type Planned PDOP Air temperature Relative
humidity Atmospheric pressure Solar wind
speed Weather data were derived from the Weather
Channel and Weather Underground. Solar wind
speed data were acquired from the Advanced
Composition Explorer (ACE) Science Center at
Cal Tech.
7th Southern Forestry and Natural Resources GIS
Conference
9
Results Samples were taken on 289 days between
September 15, 2008 and September 14,
2009. Range of data accuracy (RMSE) Young
pine stand Older pine stand Hardwood stand
Best (m) 0.5 0.2 0.8
Worst (m) 38.2 46.2 28.7
Mean (m) 11.9 6.6 7.9
Median (m) 11.5 5.8 7.3
7th Southern Forestry and Natural Resources GIS
Conference
10
Results Scatter-plot of position fixes by stand
type. Accuracy is influenced by forest type.
7th Southern Forestry and Natural Resources GIS
Conference
11
Results 5-day moving average RMSE
7th Southern Forestry and Natural Resources GIS
Conference
12
Results RMSE by relative humidity - no
correlation
7th Southern Forestry and Natural Resources GIS
Conference
13
Results RMSE by atmospheric pressure - no
correlation
7th Southern Forestry and Natural Resources GIS
Conference
14
Results RMSE by air temperature - no correlation
7th Southern Forestry and Natural Resources GIS
Conference
15
Results RMSE by solar wind speed - no correlation
7th Southern Forestry and Natural Resources GIS
Conference
16
Results RMSE by planned PDOP - no correlation
7th Southern Forestry and Natural Resources GIS
Conference
17
Results Significant differences Several
significant differences between stand
types. Fall / Winter 1. Error in hardwood stand
vs. error in the younger pine stand. 2. Error in
older pine stand vs. error in the younger pine
stand. Spring / Summer 1. Error in hardwood
stand vs. error in the older pine stand. 2. Error
in hardwood stand vs. error in the younger pine
stand. 3. Error in older pine stand vs. error in
the younger pine stand. Entire year 1. Error in
hardwood stand vs. error in the older pine
stand. 2. Error in hardwood stand vs. error in
the younger pine stand. 3. Error in older pine
stand vs. error in the younger pine stand.
7th Southern Forestry and Natural Resources GIS
Conference
18
Results Significant differences No
significant difference, for data collected within
a single stand type, between seasons of the
year.
7th Southern Forestry and Natural Resources GIS
Conference
19
Discussion Results are preliminary, and more
analysis is forthcoming.
7th Southern Forestry and Natural Resources GIS
Conference
20
Questions and Comments Welcome
One year's experience with a recreation-grade GPS
receiver
Garmin Oregon 300 tested on the UGA GPS Test
Course
7th Southern Forestry and Natural Resources GIS
Conference