"\ r "\ tsHL GEORGIA FOREST RESEARCH PAPER 46 Nov., 1983 \. J V. WEATHER CONDITIONS ASSOCIATED WITH UNDERSTORY PRESCRIBED BURNING IN SOUTH GEORGIA by James T. Paul, Curtis S. Barnes and James C. Turner, Jr. Received V r MAY 3 1988 RESEARCH DIVISION DOCUMENTS UGA LIBRARIES GEORGIA FORESTRY COMMISSION The Authors JAMES T. PAUL is Research Meteorolo- gist and Project leader with the Southeastern Forest Experiment Station, Ma- con, Georgia. He received BS (1960), MS (1966), and PhD (1973) degrees from the University of Georgia. He was an Air Force meteorology student at the University of Texas from 1960-61 and became aviation forecaster, Hunter Air Force Base, Savannah, Georgia, 1961-63. He began work with the Southern Fire Labora- tory, Macon, in 1967. From 1973-75, he was head of Project THEO, a joint Forest Service-Navy research program on fog and smoke. As Project Leader of Forestry Weather Data Systems, he is responsible for the design, implementation, and operation of the Forestry Weather Interpretations System (FWIS). CURTIS S. BARNES, formerly Associate Chief of Forest Protection of the Georgia Forestry Commission, was a graduate of the University of Georgia. He joined the Commission in 1949 as ranger of Dodge County, was made Assistant District Forester of Americus District in 1950, and District Forester of Newnan District in 1951. He was transferred to the Macon headquarters in 1955 to assume the position of Assistant Fire Chief, named Regional Forester in 1960, and was Associate Chief of Forest Protection from 1969 until his retirement in December 1981. JAMES C. TURNER, JR. is retired Chief of Forest Protection, Georgia Forestry Commission, Macon. Since joining the Commission in 1947, he had held a number of positions in various divisions. He received his Bachelor of Science in Forestry from the University of Georgia in 1947. ACKNOWLEDGMENT This work was sponsored by the Georgia Forestry Commission and could not have been completed without their personnel and financial contribution. WEATHER The Uncontrollable Limiting Factor in a Prescribed Burning Program by James T. Paul, Curtis S. Barnes and James C. Turner, Jr. Fire provides the southern forester with an economical and effective tool for use in wildlife habitat enhancement, removal of hazardous accumulations of fuels, disease control, seedbed preparation, and species management. Almost 500,000 acres were burned in Georgia during 1972 for these varied management objectives (Hough and Turner 1974). Timing of a burn (hour of ignition, sea- son), firing technique (headfire, backfire, etc.), manpower and equipment, placement of firelines, and weather are elements a forest manager must consider for a successful burn. All these elements, except weather, can be partially or wholly manipulated by the forester. Consequently, weather usually is the uncontrollable limiting factor in a prescribed burning program. Typically a forester plans the burn (selects firing technique, plows fireline, etc.), then waits for the unique set of weather variables suitable for conducting his burn to meet a specific management objective. Sometimes it is a long vigil. Meteorology has progressed in this country from occasional observation and comment by medical doctors, a military expedition, or curious citizens to a hightechnology science involving electronic sensors, high-speed digital computers, satellite observations, and mathematical models used to estimate the state of the atmosphere out to 72 hours. Curiously, the use of weather data in planning and executing prescribed burning has changed relatively little in the past 20 to 40 years. Perhaps this is partially due to an incomplete knowledge of how the various weather elements interact with the fuel complex to generate reproducible rates of spread, flame height, residence time, and fireline intensity. Also, foresters frequent- ly do not have a clear understanding of sources of weather data, which weather elements are subject to unexpected change on the short time scales relevant to prescribed fire, and how these impor- tant weather/fire variables vary over relatavely short distances. For example, a manager might conduct a successful burn in one county, but in the adjoining county, weather conditions could produce a fire too cool or too hot. In this study, data are presented that describe spatial and diurnal variability of the weather elements important for prescribed burning, the frequency of major weather systems that are responsible for rapid change of these variables, and the application of data in the Forestry Weather Interpretations System (FWIS) to the prescribed burning problem. Other compensating factors include a higher windspeed, which dissipates heat from the fire, or a higher fuel moisture, which produces a lower intensity fire. Experienced burners frequently burn with air temperatures up to 60F with appropriate attention to the compensating factors. Conversely, a burn in a young plantation might require an air temperature of no more than 40 F to avoid damaging scorch. Air temperature is relatively easy to forecast, but a major departure from the forecast can occur if timing is off on the passage of a frontal system or when the percentage of sky covered by clouds is different from what is expected. 3.5 35 3.5 Figure 1. --Average February rainfall in inches, south Georgia, 1954 to 1963. WEATHER ELEMENTS THAT INFLUENCE PRESCRIBED BURNING The National Weather Service (NWS) provides Georgia forestry forecasts which include information on rainfall type (rain, rain shower, thunderstorm, etc.), the expected amount and duration, and the probability of occurrence for today, tonight, and tomorrow. NWS also pro- vides estimates of today's maximum, tonight's minimum, and tomorrow's maximum temperature; today's minimum, tonight's maximum, and tomorrow's minimum relative humidity; and highest wind for today, tonight, and tomorrow. may be important for prescribed burning. For example, during February (19541963), the average monthly rainfall varied almost 2 inches over south Georgia (figure 1). Averaging over a long period smooths the variability associated with individual storms; consequently, on a given day, variable fuel moisture over fairly short distances should be expected even though there was a general rain across the state. The exact timing of a rainfall event is difficult and is probably the most com- mon error in precipitation forecasts. However, attention to the latest available forecast should avert most problems that foresters experience with precipitation timing. Rainfall The rainfall history at a burn site is the single most important weather element influencing total fuel moisture. Lack of rainfall, especially over long periods, or recent heavy rains may result in fire intensity inconsistent with the management objective. Rainfall during October through May is usually associated with frontal systems that produce a relatively uniform spatial distribution when compared to spotty summertime showers. However, even the lower spatial variability of frontal rain Temperature Air temperature contributes to the rate at which fuels dry, but more directly, it influences needle scorch. In general, the higher the air temperature, the greater the scorch potential. Mobley et al. (1978) recommend an air temperature in the range of 30 - 50F. There are modifying factors that permit burning at higher temperatures with no great increase in the scorch potential. For example, when burning under a mature stand the crowns are high and heat generated by the fire has a better opportunity to dissipate. Relative humidity Next to precipitation, relative humidity is the major factor influencing fine fuel moisture. If the humidity is low, finer fuels (such as the upper layer of pine needles and grass) will burn within a few hours after rain. Since the finer fuels are largely responsible for rate of spread, it is of obvious importance to the manager. Rapid changes in humidity may occur with frontal passage (cold, warm, occluded, or sea breeze) and when the air becomes unstable with resultant vertical mixing. Relative humidity is also relatively easy to forecast with major departures (forecast vs. observed) occurring with timing and intensity errors on frontal systems. The guideline for relative humidity is 30-50 percent (Mobley et al. 1978). Burning at less than 30 percent is risky because of the higher fire intensity and potential for spotting usually associated with these lower humidities. Burning at humidities appreciably higher than 50 percent may result in a missed management objective. Wind Once the fire is ignited, wind has the potential to create more problems than any other weather element. Wind at a specific time and place is relatively difficult to forecast. Accuracy of wind-direc- tion forecasts increases with increasing windspeed. Less than 5-7 mph (measured at 20 feet in the open) surface friction and unequal surface heating of different soils and vegetation types interact to increase the variability of wind direction over a short period. At the low speeds, a predominant direction can be specified, but it should be understood that observed wind direction may visit all points of the compass within a 5- to 10-minute period. Many burns are planned as either headfires or backfires, and a deviation of more 1 than 45 degrees in wind direction may defeat the management objective through its influence on fire intensity, rate of spread, and the increased potential for control problems. An example of a wind- direction shift is shown in figures 2 and 3 for an experimental burn near Waycross, Georgia, during the spring of 1982. Figure 2 is typical of a backfire where the wind bends the flame into the burned area. Figure 3 is the same fire about 5 minutes after a wind shift occurred. Mobley et al. (1978) recommend a windspeed of 2 to 10 mph in the stand, or 5 to 18 mph in the open. Low windspeeds contribute to needle scorch since heat tends to rise vertically instead of being dissipated horizontally as with stronger winds. The most common cause for fore- cast error in wind (more than 5 to 7 mph) is missed timing on frontal passages and failure to anticipate degree of pressure system intensification. ti \ . ... Figure 2. -An experimental backfire near Waycross, Georgia, on February 10, 1982. Photo courtesy F ire Science Research Work Unit, Macon, Georgia. MAJOR WEATHER SYSTEMS AND PRESCRIBED BURNING Major changes in individual weather elements occur with the movement of weather systems. Those systems that commonly produce major changes, espe- cially those with the potential to change during the course of the burn, will be dis- cussed. A typical frontal system (figure 4) has cold, warm, and occluded fronts as boun- daries between warm, cool, and cold sec- A tor air masses. cold front is cold air moving over an area formerly occupied by warm air (the temperature drops with passage), and a warm front replaces cool air (the temperature increases with pass- age). An occluded front occurs when the faster moving cold front forces the warm air sector aloft. The actual frontal zone is not a sharp, narrow band but is typically a diffuse zone where rapid changes in weather occur. The cold frontal zone is usually 10 to 50 miles wide, while the zone associated with warm fronts may be more than 100 miles wide. Occluded fronts may have a wide band of weather, but those that pass through Georgia usu- ally are very similar to cold fronts. Warm fronts that influence Georgia weather usually move from the Gulf of Mexico, traveling in a northerly direction. Cold fronts approach Georgia from the west, northwest, north, and occasionally from the northeast. Occluded fronts usu- ally approach from the west. The follow- ing discussion will focus on cold fronts because the best burning conditions usu- ally occur after a cold frontal passage. Figure 3. --An experimental backfire near Waycross, Georgia, on February 10, 1982, after a wind shift. Photo courtesy F ire Science Research Work Unit, Macon, Georgia. OCCLUDED FRONT Cool Sector -- 1 1 ) Wind direction Southeast to Northeast -- (2) Windspeed Light to moderate -- (.i) Temperature Cool -- (41 Relative humidity High -- (5) Weather Continuous ram, drizzle, with occasional showers (6) Stability --Stable Cold Sector -- ( 1 1 Wind direction West to North -- (2) Windspeed Strong, typically gusting -- (3) Temperature Cold -- (41 Relative humidity Low (5) Weather - Clear sky -- (6) Stability Stable except where midday heating may produce low-level pockets of unstable air Warm Sector -- 1 ) Wind direction East to Southwest -- (2i Windspeed Light to moderate (ill Temperature -- Warm -- I4i Relative humidity Moderate to high -- (5) Weather Intermittent rain, drizzle, showers, occasionally organized line thunderstorms -- )Bi Stability Usually unstable, expecit during daylight hours COLD FRONT Figure 4.-Diagramatic frontal system with a description of weather commonly associated with each sector. Area inside the hatched line indicates where precipitation is most probable. 6 ~s ci ^5 kj to "* 4 Q 1 to j ^ 20 ^ 10 I 456789 I I I I I I I I I I 3 10 I I I I 13 14 DAYS BETWEEN SEA BREEZE PASSAGES Figure 10.-Probability (%) of a sea breeze frontal passage occurring within X days (1-14) given a passage on day 0. 8 that over water. Frontal systems and other weather systems mix the air over land and water, and convert the dividing line to a more diffuse zone. The magni- tude of the temperature range (about 18-20F) is less near the coast due to the moderating influence of the ocean. A second area of fairly rapid change in temperature range was found north of Valdosta. This is in the agriculture belt of Georgia, and the vegetation is mixed for- ests and agriculture crops. This suggests that the differing vegetative types over short distances at least contribute to the observed range. The vegetation between Valdosta and Brunswick is largely pine forests, and the spatial change of temper- ature range is small in this area. NWS observational stations are usually located at airports where large open spaces are common (figure 14). Forestry fire weather stations (figure 15) are typi- cally located in smaller clearings and are more likely to reflect local site condi- tions. Data from Georgia Forestry Com- mission stations in Turner, Mitchell, and Lowndes Counties (see figure 8 for loca- NWS tion) were chosen instead of sta- tions for the portion of this study relating to space differences in order to show maximum differences that might occur over relatively small distances. These data (figures 16-19) are useful to illustrate the spatial variability of weather. For example, if the 1300 e.s.t. observation at a fire weather station is: Temperature Relative humidity Windspeed 1-hour timelag fuei moisture-17 70 F 40 percent 10mph 6 percent the probability that similar values will occur within a 50-mile radius of the observing station is: Temperature (figure 16): Within -- 5 or less, 80 percent of the time Relative humidity (figure 17): Within -- 10 percent or less, 80 percent of the time 40 1 35 kj 30 io io 5 25 -J 3 20 ^ o. k 15 kj N ki 10 k*i Hi 5 ^r k. ^> _L 6 8 10 12 14 16 HOUR OF DAY (e.s.t.) 20 22 Figure 1 1 .--Number of sea breeze frontal passages at Harris Neck Wildlife Refuge by hour for October through May (Paul and Williams 1971). I00 r 1400^ 90 - / /I300 1 fc 80 _ / ki ^0 / ki k 60 l000 5ct Jo 50 J500 X \[600 >I630 \I700 \l7 30 \l745 \I800 XI900 1 1 1 1 1 1 1 1 1 1 5 10 15 20 25 30 35 40 45 50 DISTANCE FROM COASTLINE (MILES) Figure 12. -Probability of sea breeze frontal passage at inland stations given a passage at either 10 or 15 miles. Representative passage times (e.s.t.) are plotted on the curve (Williams 1973). mph, 80 percent of the time Thour timelag fuel moisture (figure 19): Within 2 to 3 percent or less, 80 percent of the time. Figures 16-19 are valid only for the area and months used to develop the curves. However, they highlight three major points important for prescribed fire that should be generally true for any compar- able-size area or time period: 1. Spatial variability of weather data is usually low for most days. -- Fuel moisture was calculated from temperature, relative humidity, and cloud cover. On some days the variability is high (temperature -- 15F, relative humidity +-- 30 percent, windspeed +-- 10 mph, and fuel moisture -- 15 percent. If a prescribed burn is conducted on a high variability day, the burn- er may experience different weath- er at the burn site from what existed at a central office. If he proceeds without an onsite observation (such as with a belt weather kit), the results of the burn could be dramatically different from that expected, or result in an escaped fire. DIURNAL VARIABILITY OF PRESCRIBED BURNING WEATHER Changes in weather over a 24-hour period at a given location usually exceed spatial variability within a 50-mile radius of an observation point. The days available for prescribed burning might be increased if the manager could tailor his burn time to take advantage of the 24hour variability in weather. The diurnal curves for temperature (figure 20), relative humidity (figure 21), windspeed (figure 22), and 1-hour timelag fuel moisture (figure 23) are based on 10 kt Figure 14.-National Weather Service instrument site at Cochran Field, Macon, Georgia. Figure 13.--Average November temperature range in degrees F, south Georgia, 1954 to 1963. Figure 15. -Georgia Forestry Commission fire weather station at Turner County, Georgia. 10 100 7-x ^80 k %^ &^&kk 60 <^j ^40 v K i CCl 5 20 kQ> k , , , -* 1 1 1 1 5 10 15 20 TEMPERATURE DIFFERENCE (F.,sign ignored) lOOr -v * 80- & k^ k5 5s 60- kQ K 40- ^1 "kt k k 20 I I I 10 20 30 RELATIVE HUMIDITY DIFFERENCE (% , sign ignored) Figure 16. --Probability of air temperature varying less than value specified on X axis within 50 miles of an observation, based on 2 years of 1300 e.s.t. observations. Figure 17. --Probability of relative humidity varying less than the value specified on the X axis within 50 miles of an observing station, based on 2 years of 1300 e.s.t. observations. I00 S! so k ^