Data

Home > Data > SPB

County Acres Source Data

Robert Huggett Jr.

Southern pine beetle outbreak status for a given county is defined as at least one SPB infestation (“spot”) per thousand acres of host type. The document describes where how the host acres were estimated.

Data were taken from Forest Inventory and Analysis surveys downloaded from the version 2.1 database.

Host acre screens (FORTYPCD = forest type code)

(FORTYPCD >= 160 and FORTYPCD < 169) or (FORTYPCD >= 400 and FORTYPCD < 410) where

160 to 168 loblolly / shortleaf pine group

400 to 409 oak / pine group

Calculation of host acres (EXPCURR = area expansion factor, CONDPROP = condition proportion). This was done by plot and aggregated to the county level.

Host Acres = EXPCURR * CONDPROP

Detail of FIA survey data by state

 Alabama

• FIA surveys: 1990 periodic, 2000 periodic, and 2003 annual (subcycles 3, 4, 5- midpoint year assumed to be 2002 for calculations)

Arkansas

• FIA surveys: 1995 periodic and 2004 annual (subcycles 2, 3, 4, 5- midpoint year assumed to be 2003 for calculations)

Florida

• FIA surveys: 1995 periodic

Georgia

• FIA Surveys: 1989 periodic, 1997 periodic, and 2003 annual (subcycles 1,2,3,4,5,7- midpoint year assumed to be 2001 for calculations)

Kentucky

• FIA Surveys: 1988 periodic and 2004 annual (subcycles 1,2,3,4,5- midpoint year assumed to be 2002 for calculations)

Louisiana

• FIA Surveys: 1991 periodic and 2003 annual (subcycles 2, 4,5- midpoint year assumed to be 2002 for calculations)

North Carolina

• FIA Surveys: 1990 periodic and 2002 periodic

South Carolina

• FIA Surveys: 1986 periodic, 1993 periodic, and 2001 annual (subcycles 1,2,3,4,5- midpoint year assumed to be 2000 for calculations)

Tennessee

• FIA Surveys: 1989 periodic, 1999 periodic, and 2003 annual (subcycles 2,3,4,5- midpoint year assumed to be 2002 for calculations)

Texas

• FIA Surveys: 1992 periodic and 2003 annual (subcycles 1,2,3,4,5- midpoint year assumed to be 2002 for calculations)

Virginia

• FIA Surveys: 1992 periodic and 2001 annual (subcycles 1,2,3,4,5- midpoint year assumed to be 2000 for calculations)

Processing- Acres

• For each county, annual estimates of host area were created from the FIA data. For years where data was not available, an interpolation technique filled in the gaps. A change in area was first calculated for two nonconsecutive years. This difference was then divided by the number of years between the nonconsecutive years to obtain an average annual change in area. For example, suppose area in 1980 is 100 and in 1983 it is 160. Assume that no data is available for 1981 or 1982. The change in area between 1980 and 1983 is 60. The average annual change in area over the three year period is 20, or 60 divided by 3. The interpolated area for 1981 is 120, which is 1980’s actual area plus the average annual change of 20 for 1981. The interpolated area for 1982 is 140, which is 1981’s interpolated area of 120 plus the average annual change of 20 for 1982. Finally, the interpolated area for 1983 is 160, which is 1982’s interpolated area of 140 plus the average annual change of 20 for 1983. Note that the interpolated area for 1983 should be equal to the actual area for that year.
• If the end of the data series did not correspond to the last year of the study, 2004, then host area was assumed constant between the last year of the series and 2004. For example, if the last year for which survey data is available is 1999, then the 1999 host area was assigned to 2000-2004.
• For each county for each year, the number of spots was divided by the number of thousand host acres to yield an estimate of spots per thousand acres. If there were no host acres in a county for which there were spots for a given year, then a default level of 1,500 acres was assigned to that county and spots per thousand acres was determined by dividing the number of spots by 1.5. In the dataset the county acres field still contains the missing value indicator of single period.
• The variable smacre describes the annual estimate of spots per thousand acres. This in turn was used to assign a discrete variable disc for each county for each year:
• if spots per thousand acres < .1, then disc = 0
• if .1 ≤ spots per thousand acres < 1 then disc = 1
• if 1 ≤ spots per thousand acres < 3 then disc = 2
• if 3 ≤ spots per thousand acres then disc = 3
• if an outbreak was reported to have occurred but no spot data was available, then disc = 9