Spring vs. summer fire regimes in central North Dakota

In a nutshell

Background

• From 2017-2020, a project was conducted to restore spatially-patchy Rx fire in Central ND
• Two burning regimes:
  • 4 x 40-acre, full-patch burns each spring
  • 4 x 20-acre, half-patch burns, first half in spring, second half in summer
• Mixed success in restoring fire:
  • All spring burns were completed in each year
  • Summer burns only completed in 2/4 years but not for lack of trying

Questions

• Are there biophysical limitations on summer fires?
  • e.g., maybe it can just be too green and/or too wet and/or too humid to burn
• Can trends in vegetation and weather explain why summer burns were successful in some years and not in others?
  • If so, maybe one can predict whether a summer burn is worth trying or not, or predicting summer wildfire risk
• Potential implications for intentions behind growing season burns vs. pre-European summer fire regimes?
  • Folks recognize pre-European fire regimes included considerable summer fire, and talk about greater variability in burn severity with growing-season fire as a biodiversity-friendly reason to burnin the summer. But what if pre-European summer burns were actually quite intense and burned severely because they actually only occurred under drought conditions?

Objectives

• Explore patterns in greenness and humidity to determine effects on fire effectiveness
• Identify anomalies and ranges of conditions conducive to prescribed fire

Remotely-sensed data

Pulled burn severity (NBR) and several vegetation indices from Sentinel-2 imagery for 2017-2020 growing seasons. Data sampled from regular gridded points within burn units

Vegetation indices:

• NDVI (Normalized Difference Vegetation Index)—Measure of vegetation greenness
• NDMI (Normalized Difference Moisture Index)—Measure of moisture content of vegetation
• MSI (Moisture Stress Index)—Another measure of moisture content sometimes used in fire risk analysis. Higher = drier

Vegetation data

These data are from four unburned pastures in an adjacent continuous grazing treatment to remove any confounding effects of prescribed burns.

Seasonal trends

Summer fires were successful in 2017 and 2020. Vegetation in both years was less green and generally more dry.

Blue shaded areas span the first and last dates of completed burns in the spring and summer.

Seasons summarized

• Greenness:
  • NDVI during all spring burns generally pretty low
  • Successful summer burn years higher greenness than any spring…
  • … still slightly but distinctly lower than unsuccessful years
• Moisture content:
  • High variability in spring seasons across years
  • Successful summer seasons drier than in unsuccessful years

Data summarized by year from shaded blue ranges above.

Burn severity

Overall it appears greener fuels consistently result in less-complete burns (although 2018 was weird).

Trends between fuel moisture content and severity were less consistent. Again 2018 behaved strangely—fuel in 4 burn units in 1 block of pastures was extremely dry but still didn’t burn with particularly high severity; less-dry fuel in the other block tended to burn slightly more severely. The dry block was burned a couple weeks earlier than the other block.

Vegetation indices pulled from closest pre-burn image, same as pre-burn NBR.

Weather data

Pending