Computer Calculation of the Keetch-Byram

Drought Index-Programmers Beware!

Martin E. Alexander'

Ph.D. scholar. Department of Forestry. Australian National University. and

visiting fire researcher. National Bushfire Research Unit. CSIRO Division
of Forestry and Forest Products. Canberra. Australian Capital Territory

The Keetch-Byram Drought Index not 0.830 (fig. I). The end result of tralia, Bureau of Meteorology,
or KBDl (Keetch and Byram (968) this error is a drought factor that is Melbourne, Victoria). it's clear that
has been or is still being used as a always slightly higher than the cor­ the errors in Equations 15 and 18
guide for estimating the cumulative rect value (table I). Crane (1982) were both typographical in nature.
moisture deficiency in deep duff or also suggested that the last constant and the constant in the former should
upper soil layers. Such information is in the numerator of Equation 15 on have been 0.2113. It's worth noting
needed for planning fire management the same page should have been that the drought factor tables con­
operations in many regions of the tained in Keetch and Byram's (1968)
world (McArthur 1966, 1967; Che­ report. which are based on Equation
ney 1971; Mount 1972; Valentine • . . there are two significant 18. are correct however.
1972; Wade and Ward 1973; Burgan, typographical errors in USDA For­ Just how insignificant are these
Fujioka, and Hirata 1974; Just 1978; est Service Research Paper SE-38 sources of error in calculating the
Noble, Bary, and Gill 1980; Crane dealing with the drought index KBDl? On a day-to-day basis, the
1982; Sirakoff 1985; Swart 1986; developed by John J. Keetch and error may have only a small effect on
Burgan 1988; Melton 1989; Don­ George M. Byram ... the mis­ the resultant value (table I). How­
aldson and Paul 1990; Jordan 1990). prints in Equations IS and 18 have ever, a computer-calculated value
As well, the KBDl has been widely been corrected in a 1988-revised would eventually depart considerably
utilized in various fire research stud­ reprinting of the publication . . . • from the correct value due to the
ies (Burgan 1976; Haines, Johnson, cumulative nature of the KBDl,
and Main 1976; Dayananda 1977; especially during a rainless period
Miller 1978; Olson 1980; Lorimer 0.213 instead of 2.113. However. in (Fujioka 1991). There will of course
and Gough 1982, 1988; Hall and a review draft of Keetch and Byram always be differences between
Gwalema 1985; Johansen 1985; Van (1968) dated October 22, 1966. equation-calculated values and those
Wagner 1985; Burrows 1987; Gill, which was kindly provided by D.R. derived from tables when it comes to
Christian, Moore, and Forrester Packham (Commonwealth of Aus- fire danger indices (Deeming 1975).
1987; Brown, Booth, and Simmer­
man 1989).
It has come to my attention (Crane
1983) again that there are two sig­
English unit equation [corrected] from Keetch and Byram (1968)
nificant typographical errors in the
original 1968-published USDA Forest dO [800 - OJ [0.968 exp (O.0486D - 8.30] ciT x -3
10
=

Service Research Paper SE-38 deal­ 1 + 10.88 exp (-o.0441R)

ing with the drought index developed
by John J. Keetch and George M.
8.1. unit equation from Crane (1982)
Byram. Crane (1982) determined that
d [203.2 - OJ [0.968 exp (0.0875T + 1.5552) - 8.30) ciT x 0-3
the equation used to calculate the O 1
=

1 + 10.88 exp (-o.OO1736R)

daily drought factor was in fact
incorrect. The last constant in the
numerator of Equation 18 on page 31 QuanIIIy English units 5.1. units

of Keetch and Byram's (1968) pub­ dO 0r0ughI factor 0.01 in nvn

Q MoisIunI deftciencyl 0.01 in nvn
lication should have been 8.30 and T Dally IIIIIlCImum temperature Of" "C
R .. Mean annual precipiIatIon in mm
liT Time increment =1 day =1 day
I The author. a fire research officer with the I yesterday's l<BOlorv8Jue as reduced by the daily net poec:ipilalloll (i.e .. the amount in excess 01 0.20 in or 5.1 nvn).
Northwest Region of Forestry Canada stationed �-:'0fi:;;:r�:'�" " ); <".;',;'Y
at the Northern Forestry Centre in Edmonton. ':����'�":
AS. is presently on professional development figure I-The twO.. veTlJions of the equation used to calculate the doily drought Jactor in computing
and educational leave in Australia. the Keetch-Byram Drought Index (KBDI).

1990 Volume 51, Number 4 23

Table l-Increase in the value of the daily drought factor of the Keetch-Byram Drought Index Crane. W.J.B. 1981. Computing grassland and
(KBDI) as a result of the t1.pographical error in Equation 18 of Keetch and Byram (19681. Please forest fire behaviour. relative humidity and
note that due 10 the nature of the error in Equation 18. the increase abOl'e the actual value is inde­ drought index by pocket calculator. Aus­
pendelll of daily maximum temperature. tralian Forestry. �5(:1): 89-97.
Crane. W.J.B. 1983. Fire danger and drought
index » warning>'. Institute of Foresters of
Mean annual Yesterday's KBDI' or value as reduced by the Australia Newslener. 24(4): 27.
precipitation daily net precipitation Dayananda. P.W.A. 1977. Stochastic models
for forest fires. Ecological Modelling. 3(4):
(in) (mm) 100 200 300 400 500 600 700 800 309-313.
10 254 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 Deeming. J.E. 1975. Calculating fire-danger
20 508 1.0 0.8 0.7 0.5 0.4 0.3 0.1 0 ratings: computer vs. tables. Fire Manage­
30 762 1.3 1.1 1.0 0.8 0.6 0.4 0.2 0 ment Notes. 36(1): 6-7.9.
40 1,016 1.8 1.6 1.3 1.0 0.8 0.5 0.3 0 Donaldson. B.G.; Paul. J.T. 1990. NFDRSPC:
50 1,270 2.4 2.0 1.7 1.4 1.0 0.7 0.3 0 The National Fire-Danger Rating System on
60 1,524 3.0 2.5 2.1 1.7 1.3 0.8 0.4 0 a personal computer. Gen. Tech. Rep. SE-
70 1,778 3.5 3.0 2.5 2.0 1.5 1.0 0.5 0 61. Asheville. NC: U.S. Department of
80 2,032 4.0 3.4 2.8 2.3 1.7 1.1 0.6 0 Agriculture. Forest Service. Southeastern
Forest Experiment Station. 49 p.
, In the Original formulation of the KBOI, 800 represented the maximum possible value. However. the metriC or S.l. unit scale
of the KBOI technically limits the value to 203. Fujioka. F.M. 1991. Starting up the Keetch­
Byram Drought Index. In: Proceedings of
the 11th conference on fire and forest mete­
To my knowledge, an errata to In: MacIver. D.C.: Auld. H.: Whitewood. orology; 1991 April 16-19. Missoula. MT.
R., eds. Proceedings of the 10th conference Bethesda. MD: Society of American For­
Keetch and Byram (1968), which
on fire and forest meteorology: 1989 April esters [in press J.
highlights the aforementioned prob­ Gill. A.M.; Christian. K.R.; Moore. P.H.R.;
17-21: Ottawa. ON. Ottawa. ON: Forestry
lems, has never been issued. Canada and Environment Canada: 406-412. Forrester. R.l. 1987. Bushfire incidence.
However, the misprints in Equations Burgan. R.E. 1976. Correlation of plant mois­ fire hazard and fuel reduction burning. Aus­
15 and 18 have been corrected in a ture in Hawaii with the Keetch-Byram tralian Journal of Ecology. 12(3): 299-306.
Drought Index. Res. Note PSW-307. Hall. J.B.; Gwalema. W.N.K. 1985. Drought
1988-revised reprinting of the origi­
Berkeley. CA: U.S. Department of Agricul­ indices and fire danger indices at Morogoro.
nal publication, although no mention
ture. Forest Service. Pacific Southwest Tanzania. Forest Ecology and Management.
of these corrections is made. This Forest and Range Experiment Station. 6 p. 10(2): 125-134.
note has been prepared to alert those, Burgan. R.E. 1988. 1988 revisions 10 the 1978 Haines. D.A.; Johnson. V.J.; Main. W.A.
who may be calculating the KBDI by National Fire-Danger Rating System. Res. 1976. An assessment of three measures of
Pap. SE-273. Asheville. NC: U.S. Depart­ long tenn moisture deficiency before critical
computer, to these two errors, since
ment of Agriculture. Forest Service. fire periods. Res. Pap. NC-131. St Paul.
it's not always readily apparent
Southeastern Forest Experiment Station. MN: U.S. Department of Agriculture. Forest
whether they have been detected by 39 p. Service. North Central Forest Experiment
other users. The corrected version of Burgan. R.E.; Fujioka. F.M.; Hirata. G.H. Station. 13 p.
Equation 18 and the one rederived by 1974. A fire danger rating system for Johansen. R.W. 1985. Effect of drought on
Hawaii. Fire Technology. 10(4): 275-281. live fuel moisture content. In: Donoghue.
Crane (1982) in terms of the Interna­
Burrows. N.D. 1987. The Soil Dryness Index L.R.; Martin. R.E . • eds. Proceedings of the
tional System (S.I.) of units are for use in fire control in the south-west of eighth conference on fire and forest mete­
presented here (fig. I) in the interest Western Australia. Tech. Pap. No. 17. orology: 1985 April 29-May 2: Detroit. MI.
of completeness. Furthermore, the Perth. WA: Western Australian Department SAF Pub!. 85-04. Bethesda. MD: Society
references compiled here constitute a of Conservation and Land Management. of American Foresters: 47-51.
37 p. Jordan. D.W. 1990. The value of drought
selected bibliography on the KBDI..
Cheney. N.P. 1971. Forest industries feasi­ index in the Country Fire Authority. In: Pro­
bility study. Zambia. Fire protection of ceedings of the third Australian fire weather
Literature Cited industrial plantations. FO:SF/ZAM 5 Tech. conference; 1989 May 18-20; Hobart. TAS.
Rep. 4. Rome. ITALY: Food and Agricul­ Melbourne. VIC: Commonwealth of Aus­
Brown. J.K.; Booth. G.D.; Simmennan. D.G. ture Organization of the United Nations. tralia. Bureau of Meteorology: 133-139.
1989. Seasonal change in live fuel moisture United Nations Development Programme. Just. T.E. 1978. Extreme fire weather in
of understory plants in western U.S. aspen. 71 p. Queensland. Tech. Pap. No. 9. Brisbane.

24 Fire Management Notes

 OLD: Queensland Department of Forestry. American Foresters: 241-245. Van Wagner. C.E. 1985. Drought. timelag.
17 p. Sirakoff. C. 1985. A correction to the equa­ and fire danger rating. In: Donoghue. L.R.;
Keetch. J.J.; Byram. G.M. 1968. A drought tions describing the McArthur forest fire Martin. R.E .• eds. Proceedings of the eighth
index for forest fire control. Res. Pap. SE- danger meter. Australian Journal of Ecol­ conference on fire and forest meteorology;
38. Asheville NC: U.S. Department of Agri­ ogy. 1(}(4): 481. 1985 April 29-May 2; Detroit. MI. SAF
culture. Forest Service. Southeastern Forest Swart. R.K. 1986. Drought index for planta­ Publ. 85-{)4. Bethesda. MD: Society of
Experiment Station. 32 p. [Revised Novem­ tions. Johannesburg, SOUTH AFRICA: American Foresters: 178-185.
ber 1988.) First Bowring Protection Consultants (Ply) Wade. D.D.; Ward. D.E. 1973. An analysis
Lorimer. C.G.; Gough. W.R. 1982. Number Limited. 8 p. of the Air Force Bomb Range Fire. Res.
of days per month of moderate and extreme Valentine. 1.M. 1972. Drought index for fire Pap. SE-\05. Asheville. NC: U.S. Depart­
drought in northeastern Wisconsin. 1864- control-a measure of seasonal severity. ment of Agriculture. Forest Service.
1979. For. Res. Note 248. Madison. WI: For. Establish. Int. Rep. No. 23. Rotorua, Southeastern Forest Experiment Station.
University of Wisconsin-Madison. Depart­ NEW ZEALAND: New Zealand Forest 38 p.
ment of Forestry. 14 p. Service, Forest Research Institute. 15 p.
Lorimer. C.G.; Gough. W.R. 1988. Frequency
of drought and severe fire weather in north­
eastern Wisconsin. Journal of Environmental
Management. 26(3): 203-219. 'The 1992 National
McArthur. A.G. 1966. The application of a Wildland Fire Trajning
drought index system to Australian fire con­
trol. Canberra. ACf: Commonwealth of
. ConferenCe,,;:; .
Australia. Department of National Develop­
ment. Forestry and Timber Bureau. Forest The conference spoilsored by the .
Research Institute. 18 p. National Wildfire Coordinating
McArthur. A.G. 1967. Fire behaviour in Group's Training Working Team
eucalypt forests. Leafl. No. \07. Canberra. every other year is scheduled to be
ACf: Commonwealth of Australia. Depart­ held in 1992 in Orlando. FL. on Feb­
ment of National Development. Forestry and
ruary 20-22. at the Clarion Plaza
Timber Bureau. 36 p.
Hotel Convention Center on Interna­
Melton. M. 1989. The Keetch/Byram Drought
tional Drive.
Index: a guide to fire conditions and sup­
The theme of the training con­
pression problems. Fire Management Notes.
5(}(4): 3{}-34. ference is "Training. Performance.
\ I \ I () 11 \ () I I (1 111 () I I (J t\
Miller. R.K. 1978. The Keetch-Byram Technology-Visions of Tomorrow."
Drought Index and three fires in upper Many varied training sessions. with
Michigan. 1976. In: Preprint Volume. 5th speakers and workshops. will high­
national conference on fire and forest mete­ light this theme. An important topic funds in the next fiscal year. For fur­
orology; 1978 March 14-16; Atlantic City. for all of us will be the new fire ther information. contact Jim Whitson.
NJ. Boston. MA: American Meteorological Florida Division of Forestry. 3125
suppression curriculum and its
Society: 63-67.
development. Conner Boulevard. Tallahassee. FL
Mount. A.B. 1972. The derivation and testing 32399-1650; telephone 904-488-
Plan now to attend this important
of a soil dryness index using run-off data.
conference. and be sure to budget 6111. •
BUll. No. 4. Hobart. TAS: Forestry Com­
mission. Tasmania. 31 p.
Noble. I.R.; Bary. G.A.V.; Gill. A.M. 1980.

WILDFIRE...
McArthur's fire-danger meters. Australian
10urnal of Ecology. 5(2): 20\-203.
Olson. C.M. 1980. An evaluation of the

Don't Let It Get

Keetch-Byram Drought Index as a predictor
of foliage moisture content in a chaparral
community. In: Martin. R.E.; Edmonds,

Too Close To Home!

R.L.; Faulkner. D.A.; and others. eds. Pro­
ceedings of the sixth conference on fire and
forest meteorology; 1980 April 22-24; Seat­
tle. WA. Washington. DC: Society of FLORIDA DEPARTMENT OF AGRICULTURE AND
CONSUMER SERVICES· DIVISION OF FORESTRY

1990 Volume 51, Number 4 25