Fig 1.
The study area and some of the physical features described in the text [15].
Table 1.
Mean (±SD) seasonal changes for 2070–2099 (30-year mean) relative to 1961–1990 (30-year mean) for the study area as simulated by five coupled atmosphere-ocean general circulation models (AOGCMs) under the A2 greenhouse gases emissions scenario.
Fig 2.
Biome assignment scheme for LPJ simulated vegetation.
(A) Simulated woody plant functional type (PFT) height (m) is used to distinguish forest (>10 m) from open forest/woodland (≤10 m). (B) Savanna/grassland/steppe is assigned when woody PFT foliar projective cover (FPC) is ≤0.30 and the fire return interval is <63 years. (C) Alpine grass/shrub is assigned when annual growing degree days (5°C base; GDD5) are ≤350 (Table 2).
Table 2.
LPJ plant functional type life form and bioclimatic limits.
Fig 3.
Evaluation of LPJ simulations.
Evaluation was done using (A) Küchler [49,50] potential natural vegetation (PNV) data and (B) Advanced Very High Resolution Radiometer (AVHRR) land cover data from April 1992 to March 1993 [51]. Both data sets were reclassified into forest, grass, and shrub categories. Data excluded from analyses included PNV and land cover vegetation types not simulated by LPJ (e.g., agriculture, urban areas, wetlands). The PNV data were compared with (C) 1901–1930 [29] LPJ simulations and the land cover data were compared with (D) 1964–1993 [29] LPJ simulations, with (E, F) areas of disagreement displayed for each comparison. The legend indicates the type of disagreement. For example, the Forest-Grass category indicates areas where the (E) PNV or (F) land cover data recorded forest vegetation but LPJ simulated grass vegetation.
Table 3.
Percent agreement of LPJ simulated vegetation with Küchler [49,50] potential natural vegetation (PNV) form.
Table 4.
Percent agreement of LPJ simulated vegetation with Advanced Very High Resolution Radiometer (AVHRR) remotely sensed land cover data [51].
Fig 4.
Vegetation was simulated for (A) 1961–1990 using CRU TS 2.1 climate data [29] and for (B-F) 2070–2099 using climate projections from CCSM3 [32], CGCM3.1(T47) [33], GISS-ER [34], MIROC3.2(medres) [35], and UKMO-HadCM3 [36] coupled atmosphere-ocean general circulation models (AOGCMs). PFT = plant functional type.
Table 5.
Simulated biomes (% of study area grid cells) for 1961–1990 and 2070–2099 (bold text indicates biomes that are simulated to increase under projected future climates).
Fig 5.
Regional examples of LPJ simulated vegetation.
LPJ 1961–1990 [29] simulated vegetation for (A) the Salmon River region of Idaho and (B) Mount Rainier, Washington. (C) Willamette Valley, Oregon, historical (1961–1990 [29]) and projected future (2070–2099) vegetation as simulated by the (D) GISS-ER [34] and (E) MIROC3.2(medres) [35] coupled atmosphere-ocean general circulation models (AOGCMs). (F) Historical (1961–1990 [29]) and (G) simulated future (2070–2099, CCSM3 [32]) vegetation for the Canadian Coast Range. PFT = plant functional type.