6.5 Preliminary Results from a Two-Year Intensive Field Study of Oklahoma Grassland Fuels

Wednesday, 6 May 2015: 11:30 AM
Great Lakes Ballroom (Crowne Plaza Minneapolis Northstar)
J. D. Carlson, Oklahoma State University, Stillwater, OK; and D. M. Engle, D. Twidwell, A. S. West, S. D. Fuhlendorf, and T. Ochsner

Herbaceous fuels play a critical role in fire behavior not only in grasslands but also in mixed fuel complexes where herbaceous fuels are a sizeable component. It is these fuels which are frequently involved in the initiation and maintenance of wildland fire. In their dead fuel phase, herbaceous fuels constitute 1-hour fuels and contribute greatly to fire danger and spread. In their live phase, herbaceous fuels can serve as a heat source or sink depending on their fuel moisture. Thus, they are a fitting subject for study, especially considering the enormous land area over which they occur worldwide. As an example, millions of acres are burned annually by wildfire and prescribed fire in the Great Plains of North America. During the 2005-2006 fire season, thousands of quick-moving grassland fires burned over 3 million acres in Texas and Oklahoma alone, while claiming 25 lives and destroying over 1100 homes.

Some of the biggest unknowns in fire danger/behavior modeling pertain to live and dead fuel loads, as well as live fuel moisture (not only what the values are but also how they change throughout the year). Methods such as those used in the fire danger model in the operational OK-FIRE system in Oklahoma are largely unverified and based on the assumption that the dynamics and values of these variables (live herbaceous and woody loads, total 1-hr dead fuel load, and live herbaceous and woody fuel moisture) are a function of overall vegetation greenness (derived from satellite remote sensing in the case of OK-FIRE). In addition, these methods have largely ignored the effect of surface weather conditions (e.g., precipitation) on fuels as well as the effect of soil conditions (soil moisture and temperature). For example, in a recent article comparing measured soil moisture with remote sensing variables, Qi et al. (2012) showed observed variations in live fuel moisture throughout the year were best correlated with soil moisture. Soil moisture networks are increasing in spatial density across the country, yet the science to support the use of such data in fuel modeling is lacking.

A large multi-year grant to study the dynamics of such fuel bed characteristics in grassland fuels was awarded to Oklahoma State University (OSU) by the Joint Fire Science Program in 2011. Using intensive biweekly field sampling data from 2012 and 2013, dynamic vegetation models are being developed for grassland fuels under variable fire and grazing regimes. Variables to be modeled include live and dead herbaceous fuel loads, live to total fuel load ratio, and live fuel moisture. A variety of input data are being used for modeling purposes: weather and soil moisture data, and remote sensing data from satellites and a hand-held spectrometer.

After some initial discussion of the importance of these fuel bed variables to fire behavior and the uncertainties in modeling systems, the presentation will provide an overview of the intensive field study of perennial grasslands which took place in 2012 and 2013 at the OSU Range Research Station near Stillwater. Three pastures consisting of spatially and temporally variable fire and grazing were involved. Within each pasture (about 1/2 mile by 1/2 mile) are six unfenced sections (patches) of approximate dimension 1/6 by 1/4 mile. Each of these patches is burned in a three-year rotation (patch burning); two of the six patches are burned in any given year (summer, winter burns). With no internal fencing, patches are grazed preferentially according to the spatial-temporal burning pattern. The two-year field study involved three patches (the summer burns) in each of the three pastures, for a total of nine patches. An Oklahoma Mesonet station adjacent to one of the pastures provides weather, soil temperature, and soil moisture data for the study. Also, earlier installed digital precipitation gauges in each of the three pastures and soil moisture/temperature sensors in each of the nine studied patches provide additional localized data.

Grassland fuel bed properties were sampled biweekly in each of these nine patches from May through December 2012 and from March through November 2013. Each sampling consisted of field weighing of six separate live fuel and dead fuel clippings as well as the overall live-dead mix taken from within a 0.25 m2 quadrat at 12 locations along a transect (there are separate transects for each of the sampling weeks). Live and dead fuel moisture were calculated (after oven drying) from the separate live and dead fuel samples, and live and dead fuel loads were estimated using the constituent differential method. In addition, other variables were measured, such as fuel bed depth, canopy reflectance in five different wavelength bands using a hand-held spectrometer, and surface-to-volume ratio. In 2013 we also conducted biweekly measurements of live foliage moisture in eastern redcedar, an invasive species in grasslands, as well as the duff moisture content underneath the cedar.

The presentation will conclude with a discussion of preliminary results from this field study. Variations over the two-year period of the following measured/calculated variables will be presented: total fine fuel load, live fuel load, dead fuel load, live to total fuel load ratio, live fuel moisture, and fuel bed depth. In addition, a variety of vegetative indices calculated from the five wavelength bands of the hand-held spectrometer will be discussed as to their possible use in modeling some of the variables listed above. These indices will include such common ones as NDVI and NDII, but also less common ones such as the moisture stress index, simple ratio index, red green ratio index, and enhanced vegetation index. Finally, soil moisture correlations to some of the above variables will be discussed.

- Indicates paper has been withdrawn from meeting
- Indicates an Award Winner