2lG
It. DAVIJENMIRE
with the distal ands igniting mostly after they topple over. While
l,in(lenmut,h and Byrani (1948) and liyram (1958) found that within at
least 25 cm of t,he ground surface the maximum temperature in backfires is slightly higher than in headfires, IVeat, (1965) states that the
reverse is usually true.
Heyward ( 1 938) measured fir(! q)nwling ncrow the grass cover of the
Atlant,ic Coctstd l'lnin a t the rats of 3-4 km/h with flames rising to
3.6 ni. In Aust,ralia firc wns observed to spread through wheat stubble
at I .2 km/h (McArthur, 1966). The rate of fire spread varies approximat,ely with tdie square of wind velocity, so that burned areas tend to
be elongate in ratio to wind velocity.
l'itot and Masson (1051 ) used thermocouples to meamre temperatures
a t elevations of 0.0, 0.5 and 1.4 m as fire spread across savanna in
Senegal. The temperature ranges for these elevations were 90-140,
285-560, and 140-375°C. No temperttt,ure over 200°C was maintained
for more than 1 min. At the soil xurface temperatures exceeding 50°C
were maintained for 1-3 min. In mi earlier experiment in Senegal
where t'he grass cover was over R meter tall and wind speed was 2-3
m/sec, teniperature :kt the ground level rose to 7lfi'C (Masson, 1949).
I n derived savanna in western Africa where the fuel cover was
appreciably heavier than in Senega,l, temperatures a t the soil surface
exceeded 538OC (upper limit unknown) even when the grass wa.s burned
early in the dry season. At t,his time of year the temperature rise above
1 m elevation was negligible, but when a fire occurred late in the dry
season and fuel moisture was at, its lowest, t,he 538°C level extended
from tJhe ground to 3 m (Hopkins, 1965).
According t,o measurements made during a dry-season headfire in
Venezuelan savanna, temperatures a t t,he summit of the overhanging
wall of flame rose t,o t 6 6 0 ° C while t,hc temperature of the soil surface
80 em directly below had not yet risen above its init.ia1 38°C level
(Varcschi, 1962). Temperature8 at. the soil surface then rose slowly as
the belt of fire passed over, with maximum soil, surface temperature
(about 90OC) being reached only as the flames along the receding edge
of the belt died down. The ignition t,emperatures of dry shoots of these
savanna grasses is a t least 129°C.
Iizumi and Iwanami (1965) measured temperatures at elevations of
2, 5, 10, 15, 20, 50 and 100 cm n~ fire spread with the wind across
Miscunthus sinmsi.~ grassland on a south-facing slope in Japan. The
highest values were 300, 520, 388, 178, 116, 90 and 67%, respectively.
On bottomland where there were 2-3 times as much litter the maximal
v ~ l u e s a t the same elevations were considerably higher: 517, 604, 669,
:).A, :$!)4, 144 and 86°C. At an elevation of 2 cm the maximal temperat lire was reached 1-2 8ec after ignition on a south-facing slope, 109 BBC
- -
It. DAVIJENMIRE
with the distal ands igniting mostly after they topple over. While
l,in(lenmut,h and Byrani (1948) and liyram (1958) found that within at
least 25 cm of t,he ground surface the maximum temperature in backfires is slightly higher than in headfires, IVeat, (1965) states that the
reverse is usually true.
Heyward ( 1 938) measured fir(! q)nwling ncrow the grass cover of the
Atlant,ic Coctstd l'lnin a t the rats of 3-4 km/h with flames rising to
3.6 ni. In Aust,ralia firc wns observed to spread through wheat stubble
at I .2 km/h (McArthur, 1966). The rate of fire spread varies approximat,ely with tdie square of wind velocity, so that burned areas tend to
be elongate in ratio to wind velocity.
l'itot and Masson (1051 ) used thermocouples to meamre temperatures
a t elevations of 0.0, 0.5 and 1.4 m as fire spread across savanna in
Senegal. The temperature ranges for these elevations were 90-140,
285-560, and 140-375°C. No temperttt,ure over 200°C was maintained
for more than 1 min. At the soil xurface temperatures exceeding 50°C
were maintained for 1-3 min. In mi earlier experiment in Senegal
where t'he grass cover was over R meter tall and wind speed was 2-3
m/sec, teniperature :kt the ground level rose to 7lfi'C (Masson, 1949).
I n derived savanna in western Africa where the fuel cover was
appreciably heavier than in Senega,l, temperatures a t the soil surface
exceeded 538OC (upper limit unknown) even when the grass wa.s burned
early in the dry season. At t,his time of year the temperature rise above
1 m elevation was negligible, but when a fire occurred late in the dry
season and fuel moisture was at, its lowest, t,he 538°C level extended
from tJhe ground to 3 m (Hopkins, 1965).
According t,o measurements made during a dry-season headfire in
Venezuelan savanna, temperatures a t t,he summit of the overhanging
wall of flame rose t,o t 6 6 0 ° C while t,hc temperature of the soil surface
80 em directly below had not yet risen above its init.ia1 38°C level
(Varcschi, 1962). Temperature8 at. the soil surface then rose slowly as
the belt of fire passed over, with maximum soil, surface temperature
(about 90OC) being reached only as the flames along the receding edge
of the belt died down. The ignition t,emperatures of dry shoots of these
savanna grasses is a t least 129°C.
Iizumi and Iwanami (1965) measured temperatures at elevations of
2, 5, 10, 15, 20, 50 and 100 cm n~ fire spread with the wind across
Miscunthus sinmsi.~ grassland on a south-facing slope in Japan. The
highest values were 300, 520, 388, 178, 116, 90 and 67%, respectively.
On bottomland where there were 2-3 times as much litter the maximal
v ~ l u e s a t the same elevations were considerably higher: 517, 604, 669,
:).A, :$!)4, 144 and 86°C. At an elevation of 2 cm the maximal temperat lire was reached 1-2 8ec after ignition on a south-facing slope, 109 BBC
- -