This presentation was delivered at the third Asia-Pacific Forestry Week 2016, in Clark Freeport Zone, Philippines. The five sub-thematic streams at APFW 2016 included: Pathways to prosperity: Future trade and markets Tackling climate change: challenges and opportunities Serving society: forestry and people New institutions, new governance Our green future: green investment and growing our natural assets

1. Climate Modeling for the Asia-Pacific
-- ClimateAP and its applications
Tongli Wang, PhD
Faculty of Forestry
University of British Columbia

2. Change in global temperature
2
IPCC fifth report

3. Historical change in CO2 level and
temperature
3Source: http://www.brighton73.freeserve.co.uk/gw/paleo/400000yearslarge.gif

4. C02 emission scenarios
4

5. Projected future climates (IPCC AR5)
5

6. Forests and climate change
• Climate change mitigation
• CO2 Sequestration
• Green materials – to reduce the use
of CO2 emission intensive materials
• Adaptation to climate change
• To serve as a sink not a source of CO2
6
Carbon sink Carbon source

7. Effects of climate on forest trees
• Climate is a major environmental factor
affecting the performance of trees
(environmental effect)
• Climate is also the most important
evolutionary force causing speciation
and within-species variation (genetic
effect) in plants
7

8. Globally, forest types are associated
with climate types

9. Regionally, each tree species is adapted to a
range of climatic conditions - climatic niche

10. Climate also shapes among-population variation
along climate gradients within a species
Species range of lodgepole pine
Climatic niche
Productivity

11. Climate change causes mismatches between
the climate that trees adapted and the climate that
trees are going to experience
11
Productivity
Current Future

12. Some existing trees will move out
their suitable climate habitat
Waste of natural
resources
12

13. Current climatic
niche distribution
Future climatic
niche distribution
Lagging edge
population
extirpation
Adaptation over
generations using
standing variation
and gene flow
Natural migration
from leading edge
Natural
population
responses
Climaticgradient
warm
cold
Aitken et al. 2015

14. A framework for climate change
adaptation in forestry
14
• Climate data
for the past
and the future
Climate
data
Impact
assessment
Adaptation
strategies
• Plant-climate
relationships
• Projections on
tree species
ranges
• Climate-based
forest resources
management
• Knowledge
transfer (web
tools)

15. Challenges in getting climate data
• Not easy to access for specific locations
• Not at a desirable resolution
• Not at an expected accuracy
• Not having enough biological relevant climate
variables
• Historical and future from different sources
• Different resolutions and variables
15

16. Climate Modeling – Climate AP
Interface Coverage

17. How does it work?
• 1. Data sources
• 2. Downscaling algorisms
• 3. Calculated and derived climate variables
• 4. Integrations and downscaling of historical and
future climate data
• 5. A user-friendly interface

18. 1. Data sources ClimateAP
• Baseline data (1961-1990 normals)
• PRISM data at 4km for China
• WorldClim (4km) for the rest
• Historical annual data
• CRU: 1901 – 201x
• Future projections
• GCM projections from IPCC AR5
18

19. 2. Downscaling of the baseline data
• A combination of
• bilinear interpolation and
• Elevation adjustment
bilinear interpolation Elevation adjustment

20. Approaches for elevation adjustment
• Polynomial functions
𝑦 = 𝑎 + 𝑏1𝑥1 + 𝑏2𝑥2 + ⋯ + 𝑐1𝑥12 + 𝑐2𝑥22 + ⋯
• Partial derivative functions based on PRISM data
that incorporated topography and expert knowledge
• Dynamic local regression
elev
MAT

 = – 0.0183 + 5.5910–4  Lat – 2.2910–6  Lat  Long
)( PRISMsite
elev
MAT
elevelev 



21. Dynamic local elevational
adjustment
21
Calculate differences
between each of the 36
unique pairs

22. 22

24. 24

25. Improvements by ClimateAP
25
Fujian pilot site

26. ClimateBC is critical for climatic mapping BEC zones
Observed Predicted
Source: Hamann & Wang. 2005. Models of climate normals for genecology and climate change
studies in BC. Agricultural and Forest Meteorology 128: 211-221

28. 3. Calculated and derived climate variables
• Calculate climate variables
• Temperatures: MAT, MWMT, MCMT, TD,
• Precipitation: MAP and MSP
• Dryness: AHM and SHM
• Radiation: MAR
• Derived climate variables
• Degree-days: DD<0°C, DD>5°C, DD<18°C
and DD>18°C
• Frost: NFFD, FFP, bFFP, eFFP
• Dryness: Eref and CMD
• Snow: PAS
Weather stations

29. 4. Integrations and downscaling of
historical and future climate data
29

30. GCM data for future climate
Resolution
at 3.75 x 3.75°
30GCM data (CanESM2 RCP4.5 2050s)

31. Downscaling approach
Future climate
data
GCM
Anomaly
Reference
period
1961-1990
Replaced by
ClimateAP
Interpolation
ClimateAP
output
+

32. It facilitates predictions for management unit
32
Current Future (2050s)
Climate
Ecosystems
UBC Research Forest

33. ClimateAP output is better than original
GCM or RCM
Tmax07 of 2011

34. ClimateAP generates climate data
for the past years (1901 – 2012)
3434

35. ClimateAP includes 19 GCMs and 41 CC scenarios
for three future periods
3535

36. 36

37. 37

38. 38

39. 39

40. ClimateWNA has been Widely used for climate
related studies
40
0
200
400
600
800
1000
1200
Number of citations
Year
This will likely to be the future
for ClimateAP!
ClimateWNA

41. Future development: adding future
annual data to ClimateAP
• To reflect annual
variability and extremes
• 2 scenarios, 3 GCMs and
90 years = 540 sets
• Programming to achieve
fast access to the time
series data

42. A framework for climate change
adaptation in forestry
42
• Climate data
for the past
and the future
Climate
data
Impact
assessment
Adaptation
strategies
• Plant-climate
relationships
• Projections on
tree species
ranges
• Climate-based
forest resources
management
• Knowledge
transfer (web
tools)

43. Conditions for ecological niche models

44. Modeling and projecting bioclimate envelopes for
BC ecosystems and tree species
• The basis:
• The relationship between the realized niche and climate
variables
• Assumption:
• The realized niche is the result of local adaptation and
interactions with other species and organisms under
recent/historical climatic conditions
• Major challenges
• Model accuracy
• Uncertainty associated with uncertain future climates
44

45. Predicting BC ecosystem
Climate data Ecosystem classifications
45

46. Improvement in model accuracy
• PCA
• Discriminant
analysis
• Machine-learning
methods
• Neural network
• Random Forest +
• Optimizations
46

47. Observed vs. predicted BC ecosystems
47
BEC Zones (observed) BEC Zones (predicted)

48. A local application: predictions of Subzones
48Kamloops Timber Supply Area (at 90m)
Predicted (current)
Mapped

49. We chose 20 scenarios to represent the range
and distribution
Predicted changes in temp. and precip. for BC by 134 climate changes
scenarios for 2050s 49

50. Projected ecological responses to six selected climate change scenarios for 2050s
A consensus projection for 2050s
50

51. Consensus predictions among the 20 selected climate
change scenarios
51

52. Model-agreement among the projections based the 20
selected climate change scenarios
52
(Wang et al. 2012)

53. Results adapted into forestry policy

54. Climate niche Models – Chinese fir
• Contraction is
the substantial
• Expansion is
limited
• Concerned!
Wang et al. 2016

55. Masson pine
Wang et al. 2016

56. Chinese pine
Wang et al. 2016

57. Blue gum
Wang et al. 2016

58. Google map based ClimateAP (http://climateap.net/ )
• Spatial
visualization
• Data access
Web-based climate tools for APFNet
Output 6
This web tool makes the spatial visualization easy and simple, and delivers the output data
to users’ figure tips through a web browser.

59. Future development: more
species and ecosystems
• More species: 10 species
• China: two larch (Larix gmelinii and L.
olgensis) species, Scots pine (Pinus
sylvestris), Yunnan Pine (Pinus
yunnanensis), black locust (Robinia
pseudoacacia), Chinese cork oak (Quercus
variabilis), aspen (Populus tremula), moso
bamboo (Phyllostachys edulis)
• Chinese Taipei: Cyclobalanopsis longinux
(syn. Quercus longinux), Lithocarpus
megalophyllus (?)

60. Available ecosystem classifications

61. Climatic modeling for ecosystems
Current
Future
Wang et al. 2012

62. CC Impacts differ among ecological zones
Sub-Boreal
Spruce zone
Interior Cedar-
Hemlock zone
Current CGCM3 A2 2080s
Contraction
Expansion
62
Wang et al. 2012

63. Associate tree species with ecosystems
63
Douglas-fir
Associate Project
Future

64. Conclusions
• ClimateAP
• uses of the best available climate data and improves them
• Adds a large number of climate variables
• Offers a n all-in-one package: historical, current and future
• User friendly interface
• It facilitates ecological modelling at both species level to study the
trend and local management unit to develop adaptive management
strategies
• Our climatic and ecological models are competitive and
accessible at your figure tips.