IMSC2016 Siegmund
Impact of temperature and precipitation extremes on flowering dates of four shrub species over Germany
J. Siegmund, M. Wiedermann, J. Donges and R. Donner IMSC 2016, Canmore, Canada
Motivation environmental factors (temperature, precipitation, …)
multiple interactions
photo.naepflin.com
fauna
flora
ecosystem stability
IMSC Canmore, 2016 [email protected]
2
Motivation climate extreme environmental factors (temperature, precipitation, …)
multiple interactions
photo.naepflin.com
fauna
flora
ecosystem stability
IMSC Canmore, 2016 [email protected]
3
Data I: Meteorology 1440 stations in Germany 1901-2010 daily mean temperature, daily precipitation sum, …
IMSC Canmore, 2016 [email protected]
Österle, H., Werner, P., and Gerstengarbe, F.: Qualitätsprüfung, Ergänzung und Homogenisierung der täglichen Datenreihen in Deutschland, 1951-2003: ein neuer Datensatz, in: 7. Deutsche Klimatagung, Klimatrends: Vergangenheit und Zukunft, 9. - 11 Oktober 2006, 2006.
4
Data II: DWD Phenology Data Set 6498 stations in Germany 1951-2013 22 fruit species, 22 crops, 37 wildlife species 49 phenological events: foliation, flowering, ripening, leaf coloring, ... IMSC Canmore, 2016 [email protected]
www.DWD.de/phenology
5
Data II: DWD Phenology Data Set
Elder
Lilac
Blackthorn Hawthorn
→ massive flowering / fruit bearing shrubs → large population in central Europe → essential for many insect and bird species
IMSC Canmore, 2016 [email protected]
Southwood (1961), Duffey et al. (1974) Atkinson and Atkinson (2002)
6
Methods I: Event Coincidence Analysis 1. Two binary event time series 2. Count „coincidences“ (K) 3. Calculate coincidence rate r=K/N Case 1: NA= NB, t = 0, DT = 1 K=3 N=8 r = 0.375 Event B causes Event A
IMSC Canmore, 2016 [email protected]
Donges, J. Schleussner, C.F., Siegmund, J. and Donner, R. (2016): Event coincident analysis for quantifying statistical interrelationships between event time series. EPJ.
Methods I: Event Coincidence Analysis 1. Two binary event time series 2. Count „coincidences“ (K) 3. Calculate coincidence rate r=K/N Case 1: NA= NB, t = 0, DT = 1 K=3 N=8 r = 0.375 Event B causes Event A
IMSC Canmore, 2016 [email protected]
Donges, J. Schleussner, C.F., Siegmund, J. and Donner, R. (2016): Event coincident analysis for quantifying statistical interrelationships between event time series. EPJ.
Methods I: Event Coincidence Analysis 1. Two binary event time series 2. Count „coincidences“ (K) 3. Calculate coincidence rate r=K/N Case 1: NA= NB, t = 0, DT = 1 K=3 N=8 r = 0.375 Event B causes Event A
IMSC Canmore, 2016 [email protected]
Donges, J. Schleussner, C.F., Siegmund, J. and Donner, R. (2016): Event coincident analysis for quantifying statistical interrelationships between event time series. EPJ.
Methods I: Event Coincidence Analysis 1. Two binary event time series 2. Count „coincidences“ (K) 3. Calculate coincidence rate r=K/N
NA= NB, t = 0, DT = 1 K=3 N=8 r = 0.375 Event B causes Event A
IMSC Canmore, 2016 [email protected]
Donges, J. Schleussner, C.F., Siegmund, J. and Donner, R. (2016): Event coincident analysis for quantifying statistical interrelationships between event time series. EPJ.
Methods I: Event Coincidence Analysis 4. Testing for significance of r Calculate probability of r, presuming that events are distributed – randomly – independently – uniformly
IMSC Canmore, 2016 [email protected]
Donges, J. Schleussner, C.F., Siegmund, J. and Donner, R. (2016): Event coincident analysis for quantifying statistical interrelationships between event time series. EPJ.
Results Very early Lilac flowering vs. very warm temperatures
- three window sizes (panels) - ca. 1600 stations (y-axis) - ca. 700 window centers (x-axis) - two significance levels (red/black)
IMSC Canmore, 2016 [email protected]
12
Results Very early Lilac flowering vs. very warm temperatures
- three window sizes (panels) - ca. 1600 stations (y-axis) - ca. 700 window centers (x-axis) - two significance lvls. (red/black)
IMSC Canmore, 2016 [email protected]
13
Results Very early/late flowering vs. very warm/cold temperatures
IMSC Canmore, 2016 [email protected]
14
Results Blackthorn flowering vs. temperature (15d window), various thresholds
IMSC Canmore, 2016 [email protected]
15
Results Very early/late flowering vs. very low/high precipitation sums
-
almost no significant coincidences
-
several different event definitions tested
-
also no “eyecatching” findings for the combination of temperature and precipitation
IMSC Canmore, 2016 [email protected]
16
Summary Extraordinary met. conditions during spring significantly coincide with very unusual flowering dates (so far mainly shown for case studies or experiments)
… but also one important period in previous autumns (so far not been shown before)
Coincidence analysis can add important information to classical approaches (correlation, …) see session AM2, Thursday 11:00
IMSC Canmore, 2016 [email protected]
17
Summary Extraordinary met. conditions during spring significantly coincide with very unusual flowering dates (so far mainly shown for case studies or experiments)
… but also one important period in previous autumns (so far not been shown before)
Coincidence analysis can add important information to classical approaches (correlation, …). See Session AM2, Thursday 11:00
IMSC Canmore, 2016 [email protected]
18
Data II: DWD Phenology Data Set
IMSC Canmore, 2016 [email protected]
19
Supplementary Slides I
IMSC Canmore, 2016 [email protected]
20
Supplementary Slides II
IMSC Canmore, 2016 [email protected]
21
Supplementary Slides IV
IMSC Canmore, 2016 [email protected]
22
J. Siegmund, M. Wiedermann, J. Donges and R. Donner IMSC 2016, Canmore, Canada
Motivation environmental factors (temperature, precipitation, …)
multiple interactions
photo.naepflin.com
fauna
flora
ecosystem stability
IMSC Canmore, 2016 [email protected]
2
Motivation climate extreme environmental factors (temperature, precipitation, …)
multiple interactions
photo.naepflin.com
fauna
flora
ecosystem stability
IMSC Canmore, 2016 [email protected]
3
Data I: Meteorology 1440 stations in Germany 1901-2010 daily mean temperature, daily precipitation sum, …
IMSC Canmore, 2016 [email protected]
Österle, H., Werner, P., and Gerstengarbe, F.: Qualitätsprüfung, Ergänzung und Homogenisierung der täglichen Datenreihen in Deutschland, 1951-2003: ein neuer Datensatz, in: 7. Deutsche Klimatagung, Klimatrends: Vergangenheit und Zukunft, 9. - 11 Oktober 2006, 2006.
4
Data II: DWD Phenology Data Set 6498 stations in Germany 1951-2013 22 fruit species, 22 crops, 37 wildlife species 49 phenological events: foliation, flowering, ripening, leaf coloring, ... IMSC Canmore, 2016 [email protected]
www.DWD.de/phenology
5
Data II: DWD Phenology Data Set
Elder
Lilac
Blackthorn Hawthorn
→ massive flowering / fruit bearing shrubs → large population in central Europe → essential for many insect and bird species
IMSC Canmore, 2016 [email protected]
Southwood (1961), Duffey et al. (1974) Atkinson and Atkinson (2002)
6
Methods I: Event Coincidence Analysis 1. Two binary event time series 2. Count „coincidences“ (K) 3. Calculate coincidence rate r=K/N Case 1: NA= NB, t = 0, DT = 1 K=3 N=8 r = 0.375 Event B causes Event A
IMSC Canmore, 2016 [email protected]
Donges, J. Schleussner, C.F., Siegmund, J. and Donner, R. (2016): Event coincident analysis for quantifying statistical interrelationships between event time series. EPJ.
Methods I: Event Coincidence Analysis 1. Two binary event time series 2. Count „coincidences“ (K) 3. Calculate coincidence rate r=K/N Case 1: NA= NB, t = 0, DT = 1 K=3 N=8 r = 0.375 Event B causes Event A
IMSC Canmore, 2016 [email protected]
Donges, J. Schleussner, C.F., Siegmund, J. and Donner, R. (2016): Event coincident analysis for quantifying statistical interrelationships between event time series. EPJ.
Methods I: Event Coincidence Analysis 1. Two binary event time series 2. Count „coincidences“ (K) 3. Calculate coincidence rate r=K/N Case 1: NA= NB, t = 0, DT = 1 K=3 N=8 r = 0.375 Event B causes Event A
IMSC Canmore, 2016 [email protected]
Donges, J. Schleussner, C.F., Siegmund, J. and Donner, R. (2016): Event coincident analysis for quantifying statistical interrelationships between event time series. EPJ.
Methods I: Event Coincidence Analysis 1. Two binary event time series 2. Count „coincidences“ (K) 3. Calculate coincidence rate r=K/N
NA= NB, t = 0, DT = 1 K=3 N=8 r = 0.375 Event B causes Event A
IMSC Canmore, 2016 [email protected]
Donges, J. Schleussner, C.F., Siegmund, J. and Donner, R. (2016): Event coincident analysis for quantifying statistical interrelationships between event time series. EPJ.
Methods I: Event Coincidence Analysis 4. Testing for significance of r Calculate probability of r, presuming that events are distributed – randomly – independently – uniformly
IMSC Canmore, 2016 [email protected]
Donges, J. Schleussner, C.F., Siegmund, J. and Donner, R. (2016): Event coincident analysis for quantifying statistical interrelationships between event time series. EPJ.
Results Very early Lilac flowering vs. very warm temperatures
- three window sizes (panels) - ca. 1600 stations (y-axis) - ca. 700 window centers (x-axis) - two significance levels (red/black)
IMSC Canmore, 2016 [email protected]
12
Results Very early Lilac flowering vs. very warm temperatures
- three window sizes (panels) - ca. 1600 stations (y-axis) - ca. 700 window centers (x-axis) - two significance lvls. (red/black)
IMSC Canmore, 2016 [email protected]
13
Results Very early/late flowering vs. very warm/cold temperatures
IMSC Canmore, 2016 [email protected]
14
Results Blackthorn flowering vs. temperature (15d window), various thresholds
IMSC Canmore, 2016 [email protected]
15
Results Very early/late flowering vs. very low/high precipitation sums
-
almost no significant coincidences
-
several different event definitions tested
-
also no “eyecatching” findings for the combination of temperature and precipitation
IMSC Canmore, 2016 [email protected]
16
Summary Extraordinary met. conditions during spring significantly coincide with very unusual flowering dates (so far mainly shown for case studies or experiments)
… but also one important period in previous autumns (so far not been shown before)
Coincidence analysis can add important information to classical approaches (correlation, …) see session AM2, Thursday 11:00
IMSC Canmore, 2016 [email protected]
17
Summary Extraordinary met. conditions during spring significantly coincide with very unusual flowering dates (so far mainly shown for case studies or experiments)
… but also one important period in previous autumns (so far not been shown before)
Coincidence analysis can add important information to classical approaches (correlation, …). See Session AM2, Thursday 11:00
IMSC Canmore, 2016 [email protected]
18
Data II: DWD Phenology Data Set
IMSC Canmore, 2016 [email protected]
19
Supplementary Slides I
IMSC Canmore, 2016 [email protected]
20
Supplementary Slides II
IMSC Canmore, 2016 [email protected]
21
Supplementary Slides IV
IMSC Canmore, 2016 [email protected]
22
