South Carolina Forestry Commission

1
South Carolina Forestry Commission

• December 12, 2005
• Smoke Simulation Analysis
• Gary L. Achtemeier
• USDA Forest Service
• Athens, GA

2
Task Description

• The following set of slides summarize a
  PB-Piedmont post-burn analysis of smoke movement
  in the aftermath of a 37 acre prescribed burn
  conducted by the South Carolina Forestry
  Commission. The burn was conducted on 12 December
  2005. The model simulations run from 1600 EST 12
  December through 0800 EST 13 December.

3
Weather

• Weather during the burn and the post-burn smoke
  movement period was mostly cloudy early then
  becoming partly cloudy until 900 PM and being
  clear after that. A low pressure system over
  northeast South Carolina pushed southeast off the
  coast leaving the burn site in a broad area of
  northwesterly flow. Then, late in the analysis
  period, a secondary low center developed in
  northwest South Carolina. This system weakened
  the large scale flow allowing for drainage flows
  to dominate the local winds. The geostrophic wind
  speed, a measure of the strength of the forcing
  by the large scale (synoptic) wind field
  decreased from a moderately strong 8.9 m/sec at
  800 PM (favoring ventilation) to a weak 2.9
  m/sec at 500 AM (favoring drainage wind
  formation). As the main ridge/valley system was
  oriented NE-SW, drainage flows were opposed by
  the large scale forcing. Smoke in PB-Piedmont was
  forced up-valley early then released to flow
  down-valley later as the synoptic forcing
  weakened.

4
Modeling Information

• PB-Piedmont Version 5.0-2006
• Grid spacing 60m
• Domain size 273x243 -gt 16.4x14.6 km
• Visual domain 8.6x6.7 km
• Weather data Hourly METAR
• Elevation data USGS 30 m DEM
• Elevation contour interval 10 meters
• Elevation range is 78 198 meters

5
The full 16.4x14.6 km model domain captures most
of the branched drainage system north of the burn
site (white square). The domain also includes
most of a drainage farther west and part of a
drainage to the distant south of the burn site.
6
The PB-Piedmont viewing domain and observation
points for 12 December 2005. The white polygon
encloses the burn site. Note that observations
5,6,7 are taken in the evening and observations
8,9,10 are taken the next morning.
7
The modeled location of smoke on the ground
assumes smoldering is coming from random
locations within the model burn area. Initial
winds are from the northwest at 3.1 m/sec
beginning at 1600 EST. The time is 1800 EST, 49
minutes after sunset.
8
Smoke Location EvaluationInterpreting the
Observations Shown in the Next Slide

• Red symbol means smoke observed
• Green symbol means smoke not observed
• Circle means model matched observation
• Triangle means model unsuccessful

9
PB-Piedmont observation points 2000 EST 12
December 2005
5 No smoke (2000)
6 Light smoke (2015)
7 No Smoke (2030)
Click left mouse button to view observation
points.
10
Summary Points 5-7

• PB-Piedmont correctly placed smoke at Point 6 and
  did not place smoke at Point 5 although smoke was
  modeled in nearby drainages .
• PB-Piedmont placed scattered smoke at Point 7.
  Given the distance from the burn site to Point 7
  (3.8 km) and that only 37 acres were burned, what
  smoke that arrived at Point 7 may have been so
  diffuse as not to be observable.

11
PB-Piedmont smoke simulation at 0200 EST 13
December 2005. Strong synoptic forcing has
overcome drainage flows in the main stream basin
past Point 8 and pushed smoke up-valley to the
north and northeast of the burn site. Red dots
locate where fog (RH100) and smoke may coexist.
12
PB-Piedmont simulated smoke at 0400 EST 13
December 2005. The Synoptic forcing has weakened
sufficiently for drainage flows to overcome it.
Smoke surges southwestward past Point 8 as
drainage flows push down-valley.
13
Smoke Location EvaluationInterpreting the
Observations Shown in the Next Slide

• Red symbol means smoke observed
• Green symbol means smoke not observed
• Circle means model matched observation
• Triangle means model unsuccessful

14
PB-Piedmont observation points 0700 EST 13
December 2005
8 Light smoke in basin (0730)
9 No Smoke (0740)
10 No smoke (0745)
Click left mouse button to view observation
points.
15
Summary Points 8-10

• PB-Piedmont correctly placed smoke in the nearby
  basin at Point 8 and did not place smoke at
  Points 9 10 although smoke was modeled in
  drainages nearby Point 9.

16
Summary

• PB-Piedmont correctly placed smoke or no smoke at
  five of the six observation sites.
• Given the distance of Point 7 from the burn site
  (3.8 km) and given that only 37 acres were
  burned, it is problematic that smoke at Point 7
  was sufficient in concentration to be observed.

17
Smoke Matrix
Smoke Prediction
Yes No
0
2
No Yes
Smoke Observation
3
1