R a p i d E c o r e g i o n a l A s s e s s m e n t Rapid Ecoregional Assessment Climate and related factors: preliminary results Yukon Lowlands-Kuskokwim Mountains-Lime Hills Rapid Ecoregional Assessment Project, Alaska
Feb 24, 2016
Rapid Ecoregional Assessment
Rapid Ecoregional Assessment
Climate and related factors: preliminary results
Yukon Lowlands-Kuskokwim Mountains-Lime Hills Rapid Ecoregional Assessment Project, Alaska
Rapid Ecoregional Assessment
Schematic of MQs pertaining to climate trends
What are the projected monthly, seasonal, and annual temperature, precipitation, and length of warm and cold seasons for the REA, and how do these projections vary across time, across the region, and across varying global greenhouse gas emissions scenarios?
Where will climate change impact CEs, including subsistence species?
Rapid Ecoregional Assessment
What is SNAP?
The Scenarios Network for Alaska and Arctic Planning is a collaborative network of the University of Alaska, state, federal, and local agencies, NGOs, and industry partners.
Its mission is to provide timely access to scenarios of future conditions in Alaska and the Arctic for more effective planning by decision-makers, communities, and industry.
Rapid Ecoregional Assessment
Measuring and modeling change
Global Circulation Models (GCMs) Complex coupled models
created by national and international labs
Interactions of oceans, atmosphere, and radiation balance
Calculated which 5 of 15 models were most accurate in the far north A1B, B1 and A2 emissions
scenarios Temperature and precipitation
projections by month to 2100
Rapid Ecoregional Assessment
GCM output (ECHAM5) 2.5 x 2.5 degrees
Downscaling
Baseline values = PRISM mean monthly precipitation and temperature, 771m, 1971-2000
Adjusted and interpolated GCM outputs to historical baseline
Effectively removed model biases while scaling down the GCM projections
Frankenberg et al., Science, Sept. 11, 2009
Rapid Ecoregional Assessment
Climate Model Selection A composite (average) of all five was used, to
minimize model bias The A2 emission scenario was selected (considered
fairly probable) with some cross-comparison to A1B (more conservative).
Monthly decadal averages were used (2020s, 2050s, and 2060s), in order to reduce error due to the stochastic nature of GCM outputs
A historical baseline period of 1971-2000 was selected, to offer congruency across all SNAP-linked models.
The finest-scale (771 m) outputs were used, based on AR4 GCMs, again to provide consistency.
Rapid Ecoregional Assessment
Conceptual model of downscaled climate products
Global Circulation
Models (AR4)
Data selection
Data selection
Climate Model Biome shift
model
Permafrost Model
Fire Model
5 highest –performing
models
Monthly projected data, temp and
precip, to 2100, for 3 emission scenarios
GCM selection
Data processing
Downscaling with PRISM 1971-2000
baseline, 771m resolution
Inputs to ALFRESCO, Cliomes model, and
GIPL permafrost model; creation of freeze, thaw, and
season length interpolations
See model schematics
for full inputs
Selection of key
variables pertinent
to CEs
5-model composite, A2 and
A1B scenarios, baseline plus 2020s, 2050s,
2060s (decadal averages)
Rapid Ecoregional Assessment
Baseline climate across ecoregional landscape
Between 1949 and 1998, mean temperature increased throughout Alaska
Trends in precipitation are less clear, due to higher variability
Both temperature and precipitation varied considerably from year to year across the historical reference period
This natural variability must be taken into account when considering ongoing and future climate trends
Rapid Ecoregional Assessment
Baseline mean temperatures
Typically, the YKL ecoregion is warmest in the south in autumn, winter, and spring. However, in the summer, this pattern is reversed, with the hottest temperatures occurring to the north. This is a result of the moderating effects of the ocean and the relatively more extreme climate in interior regions.
Rapid Ecoregional Assessment
Watershed boundaries used for spatial analysis
Third-level HUCs proved to the b
Rapid Ecoregional Assessment
Baseline temperature by ecoregion
Tanana River Kvichak-Port Heiden
Upper Kuskokwim
River
Nushagak River Lower Kuskokwim
River
Central Yukon Lower Yukon Koyukuk River
-25.0
-20.0
-15.0
-10.0
-5.0
0.0
5.0
10.0
15.0
20.0
Mean temperature, 1971-2000
Mea
n te
mpe
ratu
re (°
C)
Rapid Ecoregional Assessment
Baseline precipitation
Typically, the YKL ecoregion is driest in the north in all seasons. However, precipitation varies quite widely across the ecoregion, from less than 40 mm per month to more than 170 mm. Summer rainfall is particularly variable.
Rapid Ecoregional Assessment
Baseline precipitation by ecoregion
Tanana River Kvichak-Port Heiden
Upper Kuskokwim
River
Nushagak River Lower Kuskokwim
River
Central Yukon Lower Yukon Koyukuk River0.0
10.0
20.0
30.0
40.0
50.0
60.0
70.0
80.0
90.0
100.0
Mean precipitation, 1971-2000
Mea
n pr
ecip
itatio
n (m
m)
Rapid Ecoregional Assessment
January temperature for current and three future decades, A2 scenaro (right) and A1B (below).
Projected Climate
Rapid Ecoregional Assessment
January temperature by ecoregion
Tanana River
Kvichak-Port Heiden
Upper Kuskokwim
River
Nushagak River
Lower Kuskokwim
River
Central Yukon
Lower Yukon
Koyukuk River
-25.0
-20.0
-15.0
-10.0
-5.0
0.0
Mea
n Ja
nuar
y Te
mpe
raut
re, °
C
Rapid Ecoregional Assessment
July temperature for current and three future decades, A2 scenaro (right) and A1B (below).
Projected Climate
Rapid Ecoregional Assessment
July temperature by ecoregion
Tanana River Kvichak-Port Heiden
Upper Kuskokwim
River
Nushagak River
Lower Kuskokwim
River
Central Yukon
Lower Yukon Koyukuk River
0.0
2.0
4.0
6.0
8.0
10.0
12.0
14.0
16.0
18.0
20.0
Mea
n Ju
ly te
mpe
ratu
re, °
C
Rapid Ecoregional Assessment
Summer (left) and winter precipitation for current and three future decades
Projected Climate
Rapid Ecoregional Assessment
Mean annual precipitation for current and three future decades, A2 scenaro (right) and A1B (below).
Projected Climate
Rapid Ecoregional Assessment
Mean annual precipitation by ecoregion
Tanana River
Kvichak-Port Heiden
Upper Kuskokwim
River
Nushagak River
Lower Kuskokwim
River
Central Yukon
Lower Yukon Koyukuk River
0
100
200
300
400
500
600
700
800
900
1000
Mea
n an
nual
pre
cipi
tatio
n, m
m
Rapid Ecoregional Assessment
Data at which the running mean temperature crosses the freezing point in the autumn. (Statewide context provides a range of reference).
Projected Climate
Rapid Ecoregional Assessment
Date of freeze by ecoregion
Tanana River Kvichak-Port Heiden
Upper Kuskokwim
River
Nushagak River
Lower Kuskokwim
River
Central Yukon Lower Yukon Koyukuk River16-Sep
21-Sep
26-Sep
1-Oct
6-Oct
11-Oct
16-Oct
21-Oct
26-Oct
31-Oct
5-Nov
Rapid Ecoregional Assessment
Data at which the running mean temperature crosses the freezing point in the srping. (Statewide context provides a range of reference).
Projected Climate
Rapid Ecoregional Assessment
Date of thaw by ecoregion
Tanana River Kvichak-Port Heiden
Upper Kuskokwim
River
Nushagak River
Lower Kuskokwim
River
Central Yukon Lower Yukon Koyukuk River30-Dec
19-Jan
8-Feb
28-Feb
20-Mar
9-Apr
29-Apr
19-May
Rapid Ecoregional Assessment
Permafrost: driver of change Permafrost thaw is both a result of climate
change, and a change agent in its own right In permafrost areas, the formation and
drainage of thermokarst lakes plays a key role in the hydrologic dynamics of the ecosystem
Permafrost thaw leads to multiple effects, including frost heaves, pits, gullies, differential tussock growth, localized drying, and changes in shrub and moss species abundance, productivity, and mortality
Permafrost degradation can occur in many different ways, depending on slope, soil texture, hydrology, and ice content, and each of these modes has different effects on ecosystems, human activities, infrastructure, and energy fluxes
Torre Jorgenson
Rapid Ecoregional Assessment
Permafrost Modeling Permafrost modeling was done using SNAP climate projections
as described under climate modeling, and the Geophysical Institute Permafrost Lab (GIPL) permafrost model for Alaska
Model outputs include mean annual ground temperature (MAGT) and active layer thickness (ALT)
Algorithms are dependent on the insulating properties of varying ground cover and soil types, as well as on climate variables
Resolution is 1-2km Although very fine-scale changes in micro-conditions cannot be
accurately predicted by the GIPL model, outputs provide a general picture of areas likely to undergo some degree of thaw and associated hydrologic changes
Rapid Ecoregional Assessment
Conceptual model of GIPL permafrost modeling techniques
Rapid Ecoregional Assessment
Schematic of GIPL model
Rapid Ecoregional Assessment
Schematic of MQs related to permafrost
What are the current soil thermal regime dynamics? Based on the predictions of the best available climate
models and soil temperature models, how will soil thermal regimes change in the future?
Where are predicted changes in soil thermal regimes associated with communities and transportation routes?
How and where will changes in permafrost impact vegetation?
How might changes in temperature, precipitation, evapotranspiration, and soil thermal dynamics affect general hydrology and hydrology-dependent CEs such as waterfowl in the region?
Rapid Ecoregional Assessment
Permafrost: MAGT
Mean annual ground temperature at one meter depth serves as a reasonable proxy for the presence/absence of ecologically significant permafrost.
Blue areas are frozen; white to orange areas are thawed.
Rapid Ecoregional Assessment
Tanana River Kvichak-Port Heiden
Upper Kuskokwim
River
Nushagak River
Lower Kuskokwim
River
Central Yukon
Lower Yukon Koyukuk River
-3.00
-2.00
-1.00
0.00
1.00
2.00
3.00
4.00
Grou
nd te
mpe
ratu
re a
t one
met
er d
epth
Permafrost: MAGT
Rapid Ecoregional Assessment
Permafrost: ALT
These maps depict two different variables.
In areas with permafrost (temperatures below freezing at one meter depth), the brown shades show seasonal thaw.
Blue shades show depth of winter freeze in non-permafrost areas.
Rapid Ecoregional Assessment
Predicted changes in soil thermal regime
Permafrost is expected to undergo significant thaw across much of the REA as mean annual ground temperature at one meter depth rises from below 0°C to above 0°C
Note that thaw at one meter does not equate with total permafrost loss, since deeper permafrost is likely to persist much longer, with a talik layer above it
In addition, areas that are already without permafrost are likely to experience shallower winter freezing, and areas that retain permafrost throughout the study period are likely to experience deeper summer thaw (thicker active layer)
Rapid Ecoregional Assessment
Fire Assessment and Modeling Fire is being modeled using SNAP climate
data and the ALFRESCO model in the larger context of a projected future fire regime and its effects on major vegetation classes
Climate projections, past fire history, and current vegetation patterns will be used to model patterns of fire frequency across the landscape.
Fire behavior involves stochastic elements such as the exact location of lightning strikes and the variability of weather patterns at finer time-scales than are available
Therefore, fire distribution per se will not be modeled; rather its projected average frequency across the landscape will be used to model changes in vegetation patterns and distribution
seagrant.uaf.edu
Rapid Ecoregional Assessment
MQs related to fire
What is the fire history of the ecoregion? What climatic conditions are likely to
result in significant changes to fire activity?
What is the current frequency (return interval) and the likely future frequency for fire in the ecoregion and broad sub-regions?
Rapid Ecoregional Assessment
Conceptual model of ALFRESCO fire simulation methodology
Rapid Ecoregional Assessment
ALFRESCO 1.0
Rapid Ecoregional Assessment
ALFRESCO 2.0
Rapid Ecoregional Assessment
ALFRESCO 2.0
Rapid Ecoregional Assessment
Cumulative Area Burned
Historical (1950-2011)ALFRESCO replicates
Rapid Ecoregional Assessment
Cumulative Area Burned
Historical (1950-2011)ALFRESCO replicates
Rapid Ecoregional Assessment
ALFRESCO 2.0 Alaska Frame-Based Ecosystem Code Spatially explicit state & transition model Model is driven by disturbance & climate
Historical climate data are derived from CRU Projected climate data are derived individually for the 5 best
models and the A2 emission scenario (ALFRESCO cannot use composite model because variability is too low for calibrations)
Simulates fire & vegetation succession dynamics at a 1-km spatial resolution on a 1-year time step
Transitions constrained based on observed changes Results averaged across 100 model runs per climate model = 500
model runs.
Rapid Ecoregional Assessment
Fire history and current fire regime
Fire frequency is dependent not only on the flammability of the landscape, but also on fire ignitions from lightning
Although lightning strikes are tracked by the Alaska Fire Service accuracy of measurement has been inconsistent over time, meaning that no consistent trends can be found in historical data
In some cases, climate change appears to be positively correlated with increased cloud-to-ground lightning activity
Rapid Ecoregional Assessment
Sample lightning ignitions
http://afsmaps.blm.gov/imf/imf.jsp?site=lightning
Rapid Ecoregional Assessment
Fire history from 1940 to the present(http://fire.ak.blm.gov/predsvcs/maps.php)
Rapid Ecoregional Assessment
Fire regime: ongoing effects Lichens are slow to regrow after fire Recent decades have seen marked change in tundra
ecosystems due to the interplay of climate change, wildfire, and disturbance by caribou and reindeer
Observed significant reduction of terricolous lichen ground cover and biomass
Fire can also lead to vegetation shift; in one study, it was found that shrub cover was higher on burned plots than unburned plots, and that cover of cottongrass (Eriophorum vaginatum) initially increased following the fire, and remained so for more than 14 years
Rapid Ecoregional Assessment
Tundra to Forest Transition
Adapted from Epstein et al. (2004) Journal of Biogeography
Relative importance for transitions
Rate of change Tundra/Forest
State factor controls
Climate Moderate-Fast High
Parent material Slow Low
Topography Slow Low
Disturbance Slow-Fast Medium
Environmental interactions
Permafrost/active layer Moderate-Fast High
Hydrology and snow Moderate-Fast Medium
Properties of dominant plant species
Time to dominance Slow (104 + years)
Continuity of abundance across transition Abrupt
Rapid Ecoregional Assessment
Tundra to Forest Transition
Rapid Ecoregional Assessment
Projections for Spatial Transitions
Reclassification of NALCMS Land Cover Map (2005)
Rapid Ecoregional Assessment
Tundra to Forest Transition
CCCMA ALFRESCO ReplicateBasal area of White Spruce low mid high
Rapid Ecoregional Assessment
Effects of future fire regime on major vegetation classes Fire-driven vegetative change may be at least as important as
change directly driven by temperature increases Shorter fire cycles and more frequent burning in areas that
previously saw little fire will result in an overall shift toward early-succession vegetation
Species such as willow, birch, and aspen may gain precedence over older-succession spruce in forested areas, and in tundra, faster-growing grasses may prevail over slower-growing lichens
Species that rely on early-succession vegetation (e.g. moose) are likely to gain a competitive advantage over those that require late-succession vegetation (e.g. caribou).
Habitat requirements must be examined on a species by species basis
Of particular note are habitat types that support species during times of stress or limited resources
Rapid Ecoregional Assessment
Future fire regime and aquatic resources
By 2060, rivers in the ecoregion will be in places experiencing relatively more frequent fire
Fires can add large woody debris and nutrients to rivers immediately following severe burns
Burning along river banks can exacerbate erosion
Rapid Ecoregional Assessment
Future fire regime and communities
The effects of fire go beyond the threat of losing housing and infrastructure
Downed trees and the dense brush that grows following fires cut off hunting trails
Cabins for hunter travel, hunting and trapping, (some were built on federal land long before land claims) burn down
Conversely, fire can create new browse/berry patches
Fire also has political dimensions, in that fire policy is one of many areas where local people are affected by consequences but are unable to control such policies
Rapid Ecoregional Assessment
Fire impacts on permafrost Increases in fire frequency may accelerate the thaw
of permafrost in the region, given that in areas where burns are severe and the organic layer is consumed, more rapid thaw has been observed immediately afterwards
In cases where most of the organic layer burns during an intense fire, subsequent heat transfer to the ground will be increased
Thus, estimates of permafrost thaw are likely to be conservative in areas projected to be strongly influenced by fire
Rapid Ecoregional Assessment
Climate-Biome Shift
Rapid Ecoregional Assessment
Cliomes conceptual model
Intersection between clusters and coarse-filter
CES
SNAP historical monthly temp and precip data 1971-2000 downscaled with 10-minute
CRU grids
Clustering region encompassing
Alaska and Western Canada
Clustering region encompassing
Alaska and Western Canada
18 climate-biome
clusters (“cliomes”)
Potential change in
coarse-filter CEs
PAM clustering methodology
Re-projection of cliomes into
the future
Rapid Ecoregional Assessment
Describing the clusters:growing degree days, season length, and snowfall
0
500
1000
1500
2000
2500
3000
50
70
90
110
130
150
170
190
210
230
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18G
row
ing d
egre
e da
ys
Days
abo
ve fr
eezi
ng
cluster
Days above freezing
Growing Degree Days
0
200
400
600
800
1000
1200
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18
Tota
l pre
cipi
tatio
n, m
m (
rain
wat
er e
quiv
alen
t)
Clusters
total for months with mean temperature below freezing
total for months with mean temperature above freezing
Length of above-freezing season and GDD by cluster. Days above freezing were estimated via linear interpolation between monthly mean temperatures. Growing degree days (GDD) were calculated using 0°C as a baseline.
Warm-season and cold-season precipitation by cluster. The majority of precipitation in months with mean temperatures below freezing is assumed to be snow (measured as rainwater equivalent).
Rapid Ecoregional Assessment
Projected cliomes for the five-model composite, A1B (mid-range ) climate scenario.
Alaska and the Yukon are shown at 2km resolution and NWT at 10 minute lat/long resolution .
Climate-biomeProjections
Original 18 clusters
Rapid Ecoregional Assessment
Cliome projections with REA boundary shown in black
Rapid Ecoregional Assessment
SNAP climate-biomes (cliomes) by coarse-filter CE
cliome 8
cliome 9
cliome 10
cliome 11
cliome 12
cliome 13
cliome 14
cliome 15
cliome 16
cliome 17
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
UnvegetatedSparse Vegetation/ lichenHerbaceous Lichen/ Dwarf ShrubLow Shrub/ LichenTall Shrub Spruce forest/ lichenDeciduous Forest
Rapid Ecoregional Assessment
Cliomes by time period (2 km pixels within REA boundaries)
baseline (1971-2000) 2020s 2050s 2060s0
20000000
40000000
60000000
80000000
100000000
120000000
cliome 8cliome 9cliome 10cliome 11cliome 12cliome 13cliome 14cliome 15cliome 16cliome 17cliome 18
Rapid Ecoregional Assessment
Estimated CE coverage by time period, based on cliome analysis
Deciduous F
orest
Spruce
forest/ l
ichen
Tall Sh
rub
Low Shru
b/ Lich
en
Lichen
/ Dwarf
Shrub
Herbace
ous
Spars
e Vege
tation/ li
chen
Unvegeta
ted
Unknown
0.00
0.10
0.20
0.30
0.40
0.50
0.60
baseline (1971-2000)2020s2050s2060s
Rapid Ecoregional Assessment
Projected change in coarse-filter CEs, 2010s to 2060s, based on cliome analysis
Tanana Rive
r
Kvichak-Port
Heiden
Upper Kuskokwim
River
Nushaga
k River
Lower
Kuskokwim
River
Central Y
ukon
Lower
Yukon
Koyukuk River
-20%
-15%
-10%
-5%
0%
5%
10%
Deciduous Forest Spruce forest/ lichenTall Shrub Low Shrub/ LichenLichen/ Dwarf ShrubHerbaceous Sparse Vegetation/ lichenUnvegetatedUnknown
Rapid Ecoregional Assessment
Projected change in habitat, 2010s to 2060s, based on cliome analysis
Tanana R
iver
Kvichak-Port
Heiden
Upper Kuskokwim
River
Nushaga
k River
Lower
Kuskokwim River
Centra
l Yuko
n
Lower
Yukon
Koyukuk R
iver
-20%
-15%
-10%
-5%
0%
5%
10%
15%
moose cariboubothneither
Rapid Ecoregional Assessment
Connecting fine-filter CEs and climate-related CAs: possible next steps?
Caribou: Overlay with cliomes and assess loss of lichen habitat in tabular form by hucDiscuss in relation to landcover change after fire
Swans: Overlay with summer season length maps Overlay with permafrost maps
Flycatcher: No clear connection to climate variables within the REA
Falcon:No clear connection to climate variables within the REA
Muskoxen: Overlay with coldest month data and/or winter precipitation, and discuss rain on snow events. Assess loss of lichen habitat in tabular form by huc
Beaver: Overlay with summer season length and/or permafrost maps
Wolf: Discuss in conjunction with caribou and moose habitat
Moose: Overlay with cliomes and assess increase in shrubs/forest in tabular form by huc Discuss in relation to landcover change after fire
Rapid Ecoregional Assessment
Swans: ice-free days
Swans require a minimum of 150 ice-free days to successfully fledge cygnets.
However, this limitation does not seem to play a major role in the YKL REA.
Hydrologic constraints may play a larger role than climate. However, hydrology is influenced by permafrost…
Rapid Ecoregional Assessment
Swans: permafrost
Swan habitat includes permafrost and non-permafrost areas.
However, areas projected to undergo significant permafrost thawing may be susceptible to changes in drainage that might affect habitat.