|
Titel |
A small single-nozzle rainfall simulator to measure erosion response on different burn severities in southern British Columbia, Canada |
VerfasserIn |
Ashley Covert, Peter Jordan |
Konferenz |
EGU General Assembly 2010
|
Medientyp |
Artikel
|
Sprache |
Englisch
|
Digitales Dokument |
PDF |
Erschienen |
In: GRA - Volume 12 (2010) |
Datensatznummer |
250038226
|
|
|
|
Zusammenfassung |
To study the effects of wildfire burn severity on runoff generation and soil erosion from high
intensity rainfall, we constructed an effective yet simple rainfall simulator that was
inexpensive, portable and easily operated by two people on steep, forested slopes in southern
British Columbia, Canada. The entire apparatus, including simulator, pumps, hoses,
collapsible water bladders and sample bottles, was designed to fit into a single full-sized
pick-up truck. The three-legged simulator extended to approximately 3.3 metres above
ground on steep slopes and used a single Spraying Systems 1/2HH-30WSQ nozzle which can
easily be interchanged for other sized nozzles. Rainfall characteristics were measured using a
digital camera which took images of the raindrops against a grid. Median drop
size and velocity 5 cm above ground were measured and found to be 3/4 of the
size of natural rain drops of that diameter class, and fell 7% faster than terminal
velocity.
The simulator was used for experiments on runoff and erosion on sites burned in 2007 by
two wildfires in southern British Columbia. Simulations were repeated one and two years
after the fires. Rainfall was simulated at an average rate of 67 mm hr-1 over a 1 m2 plot for
20 minutes. This rainfall rate is similar to the 100 year return period rainfall intensity for this
duration at a nearby weather station. Simulations were conducted on five replicate 1
m2 plots in each experimental unit including high burn severity, moderate burn
severity, unburned, and unburned with forest floor removed. During the simulation a
sample was collected for 30 seconds every minute, with two additional samples
until runoff ceased, resulting in 22 samples per simulation. Runoff, overland flow
coefficient, infiltration and sediment yield were compared between treatments. Additional
simulations were conducted immediately after a 2009 wildfire to test different mulch
treatments.
Typical results showed that runoff on plots with high burn severity and with forest floor
removed was similar, reaching on average a steady rate of about 60% of rainfall rate after
about 7 minutes. Runoff on unburned plots with intact forest floor was much lower, typically
less than 20% of rainfall rate. Sediment yield was greatest on plots with forest floor removed,
followed by severely burned plots. Sediment yield on unburned and moderately
burned plots was very low to zero. These results are consistent with qualitative
observations made following several extreme rainfall events on recent burns in the region. |
|
|
|
|
|