Transcript Teakettle May 2012 – mean monthly maximum temperature

m2M Conference Call
February 13, 2013
• John surface temp estimate update
• Ian update on progress with distillation of
garden array into pixel value
• Lynn update on seedling survival/recruitment
data
• NASA letter of intent/ecological forecasting
(Frank, Lorrie, Alan…)
Landsat 8!
• The following figures show the percentage of seedlings alive at the
end of the season out of the seeds that were planted versus
growing degree days at a given garden (5C threshold). These are not
corrected for germination rate.
• Apologies that this is a quick excel graph just to show this group
(curves fit to equations in excel too) , with many data points on
them (all species on one graph), and this should be considered
preliminary.
• I have several “censuses” at each site, so I can tease apart which
seedlings were alive when. Seeds->emerged seeds->survivors.
• I will try various independent variables and see which fit best for
which species (GDD, GDH, +5C, +10C, etc.)
– Lynn
Recruitment
Recruitment
Recruitment
Recruitment
Recruitment
Next steps: plant data
• First graphic: GDD vs recruitment (survival to end of
season):
– Correct for % germination (i.e. how many could we have
possibly seen)
– Growing degree days may be more relevant to survival than
emergence, and relevant period may differ among species (e.g.
pines later, oaks earlier, Teakettle has snow until June…)
– Try various degree-day calculations, and thresholds, but only for
survivorship, not emergence
• Second graphic: Survival curves:
– % first detected; % survived to second census; % survived to
final census
• Survivorship trajectories/when does mortality occur
• Apply these regression formulas to the landscape
Updates from the Davis Lab
Does interpolated surface (5 cm) temperature at 30 m/monthly resolution capture finescale temperature variability measured in the field?
• Compare average monthly max temp of all garden sensors for a given garden to
interpolated average monthly max temp for that garden
• So, comparing a single monthly interpolated value to approximately 21 sensors
x 30 days per month= >600 temp values x 6 gardens per site
Slides shown on m2M monthly conference call: 2-13-13 by Ian
Teakettle May 2012 – mean monthly maximum temperature
50.00
45.00
40.00
Deg C
35.00
Est_avg_max_5cm
Garden_avg_max_5cm
30.00
25.00
20.00
15.00
sm1
sm2
sm3
sm4
sm5
sm6
Blue=interpolated average max daily temp (Campbell method, performed by John Dingman) for given month at 5 cm
Red=average max daily temp of all garden sensors for given month at 5 cm
Bars indicate standard deviation for the whole month, not for a given day. Therefore, long bars indicate there is wide temp variability throughout
a month that isn’t necessarily evident in a single mean monthly value. The biological relevance of this is another question.
Teakettle June 2012 – mean monthly maximum temperature
50.00
45.00
Deg C
40.00
Est_avg_max_5cm
35.00
Garden_avg_max_5cm
30.00
25.00
20.00
sm1
sm2
sm3
sm4
sm5
sm6
Teakettle July 2012 – mean monthly maximum temperature
50.00
45.00
Deg C
40.00
Est_avg_max_5cm
Garden_avg_max_5cm
35.00
30.00
25.00
sm1
sm2
sm3
sm4
sm5
Arrows indicate a pattern of south slopes underpredicting max temp of full garden array
sm6
Teakettle Aug 2012 – mean monthly maximum temperature
50.00
45.00
Deg C
40.00
Est_avg_max_5cm
Garden_avg_max_5cm
35.00
30.00
25.00
sm1
sm2
sm3
sm4
sm5
sm6
Teakettle Sep 2012 – mean monthly maximum temperature
50.00
45.00
Deg C
40.00
Est_avg_max_5cm
Garden_avg_max_5cm
35.00
30.00
25.00
sm1
sm2
sm3
sm4
sm5
sm6
Discussion
• So, the surface temperature interpolation with a monthly time-step does not necessarily
capture very fine-scale variability in surface temperature
o May do so for north slopes and cold air pooling sites
o Seems to under-predict the amount of warmth on south slopes during hot summer
months
o Duration of temperature extremes are not represented in mean data...how much does
this matter from a biological standpoint?
•
Given that minimum temperature varies less over small areas based on our garden data, I
would expect that the interpolation method would show better agreement with average
minimum temperature measured at the gardens