Notes: This
is an image from several different NEXRAD locations showing rainfall intensity
at one moment in time with red symbolizing high precipitation and blue as low.
Precipitation totals can be determined by studying these images over a period
of time.
Notes: The Burd Run Watershed is located in Cumberland
County in its southwest corner. It is a small scale region that can provide for
a well-detailed study area. It has a diverse landscape with ranging topography,
vegetation, and land use.
Notes: This is a large scale image showing the
Burd Run Watershed and its location compared to the two closest NEXRAD locations--State
College, PA and Washington D.C. The study area is approximately the same distance
from each radar site.
Notes: This is a topographic map of the Burd
Run Watershed showing the variations in elevation. Rain gauges were located at
all elevations to see if there were any trends in NEXRAD innaccuracies as a result
of topography. For a full size topographic map, CLICK
HERE.
Notes: View of the mountainous
region within the study area.
Notes: This is the distribution
of land use in the watershed. The southern tip is the state forest shown in green
and the northern half is mainly agricultural land shown in orange. For a full
size land use map, CLICK
HERE
Notes:
This is a view from the mountains overlooking the Cumberland Valley.
Notes: Rain guages were installed
throughout the watershed in locations varying in elevation and land use.
Notes: This is a typical
rain gauge location in the mountains. Gauges were installed in locations with
open aras in an attempt to prevent overhead tree cover from intercepting precipitation
before it could be caught by the gauge.
Notes:
Gauges were placed along mountain roads there were easily accessible to vehicles.
Notes:
Every rain event, large or small, rain gauges were checked and the amount of precipitation
recorded. Even after small events, the guages had to be emptied to prevent inaccurate
readings after the next rain event. Notice the open field location of this guage.
Note: all gauges used were identical.
Notes: Sometimes certain
things were unforeseen, such as the seasonal growth of corn at this site, same
site as prevous slide. The corn stalks could have shielded rain from entering
the gauge.
Notes: This map shows
the location of each rain gauge in the watershed. Thiessen polygons surrounding
each site were mapped with GIS, which show the relative areas closest to the gauge.
Two points were used in the analysis that are not within the watershed.
Notes: This is an example
of the NEXRAD images that were downloaded from the internet. This particular one
is from State College showing rainfall amounts in ranges. The lighter blue shows
are that received 0-0.3 inches of rain, the darker blue show between 3/10 to 6/10,
and the green area between 6/10 to 1inch. The area in the red square is the study
area.
Notes: This is a zoomed
in image from the previous slide. We can see the distribution of rain amounts
pixel by pixel. The map of the watershed was overlaid using GIS to show the range
each gauge was part of. The pink represents the county line.
Notes: After analyzing each rain event over 3/10
inches, we found interesting results. This graph shows all of the State College
and Wahsington D.C. NEXRAD estimates combined to compare with the measure values
from the gauges. The blue shows that when the radar was inaccurate it was typically
underestimated. This could cause a problem in the case of predicting flash flooding
in a particular area.
Notes: These graphs show
the difference of rain events between 3/10 to 1 inch and the events with rainfall
over 1 inch. Larger rain events prove to be more inaccurate than accurate. The
radar understimated the actual amount of rainfall 52% of the time.
Notes: This is a comparison
between the two radar sites that are individually compared to the measured values.
State College precpitation values understimate the rain gauge for 51% of the time,
while Washington D.C. radar estimates are more balanced in its inaccuracies. Still,
it is accurate only 53% of the time.
Notes: These maps of
the watershed compare the accuracy percentage of each NEXRAD location. The red
represents a low percent accuracy, while the gray areas are of high precent accuracy.
The State colloege radar appears to be much less accurate in the mountainous portion
of the watershed. The Washington D.C. radar does not seem to have a pattern of
accuracy.
Notes: These maps are
of an example of an accurate NEXRAD estimate, in this case from State College.
On this rain event, all of the guage measurements were within the range of the
NEXRAD estimates.
Notes:
This is an example of an inaccurate NEXRAD estimate taken from Washington D.C.
