A Site Atmospheric State Best Estimate of Temperature for ...€¦ · A Site Atmospheric State Best Estimate of Temperature for Lauder, New Zealand Jordis Tradowsky1;2;3, Greg Bodeker1,

A Site Atmospheric State Best Estimateof Temperature for Lauder, New Zealand

Jordis Tradowsky1,2,3, Greg Bodeker1, Richard Querel2,Peter Builtjes3, Jurgen Fischer3

1Bodeker Scientific2National Institute of Water and Atmospheric Research (NZ)

3Freie Universitat Berlin

[email protected]

April 25, 2018

Site AtmosphericState Best

Estimate forLauder

JordisTradowsky

Introduction

Methodology

Componentsof the SASBE

Example ofthe SASBE

Summary

References

2/16

Site Atmospheric State Best Estimates

Site Atmospheric State Best Estimates (SASBEs)

I Combine measurements from multiple instruments tocreate the best-possible vertically resolved time series oftarget parameter above one site

I Contain all available knowledge about the state of targetvariable at that site

I Include an estimate of the uncertainty on every value


Estimate forLauder

JordisTradowsky

Introduction

Methodology


Example ofthe SASBE

Summary

References

2/16







Estimate forLauder

JordisTradowsky

Introduction

Methodology


Example ofthe SASBE

Summary

References

2/16







Estimate forLauder

JordisTradowsky

Introduction

Methodology


Example ofthe SASBE

Summary

References

3/16

Temperature SASBE for Lauder

I Temperature SASBE for the distributed GRUAN site atLauder and Invercargill has been published

I https://doi.org/10.5281/zenodo.1195779 andTradowsky et al. (2018)

I Available for 1997-2012

I Hourly temporal resolution

I Vertically resolved on standard pressure levels1

1ground, 925 hPa, 850 hPa, 700 hPa, 500 hPa, 400 hPa, 300 hPa,200 hPa, 150 hPa, 100 hPa, 70 hPa, 50 hPa, 30 hPa, 20 hPa, 10 hPa

https://doi.org/10.5281/zenodo.1195779


Estimate forLauder

JordisTradowsky

Introduction

Methodology


Example ofthe SASBE

Summary

References

3/16










Estimate forLauder

JordisTradowsky

Introduction

Methodology


Example ofthe SASBE

Summary

References

3/16










Estimate forLauder

JordisTradowsky

Introduction

Methodology


Example ofthe SASBE

Summary

References

3/16










Estimate forLauder

JordisTradowsky

Introduction

Methodology


Example ofthe SASBE

Summary

References

3/16










Estimate forLauder

JordisTradowsky

Introduction

Methodology


Example ofthe SASBE

Summary

References

4/16

Schematic explanation of temperatureSASBE for Lauder

Time

Δ 𝑡𝑜Anomalies Lauder

𝒕𝑺𝑨𝑺𝑩𝑬

𝐓SASBE(𝒕𝑺𝑨𝑺𝑩𝑬) =

ϕ= autocorrelation(Δ 𝑡0)

Transfer function

Weighting

Invercargill

Lauder

Diurnal cycle

Lauder component

Invercargillcomponent

180km

+ +

Anomalies Invercargill

Weighting

Invercargillcomponent

Lauder component

Diurnal cycle

Dis

tanc

e


Estimate forLauder

JordisTradowsky

Introduction

Methodology


Example ofthe SASBE

Summary

References

5/16

Calculating the upper-air diurnaltemperature cycle

I Diurnal cycle calculated by fitting Fourier series to ERA5→ climatological mean diurnal temperature cycle for everyday

I The fitting uncertainty is calculated as:

σfit =

√√√√ 81∑i=1

σ2ζi

(∂TDiur

∂ζi

)2

(1)

→ Indicates how good the regression model fits the 8years of reanalysis

I Representativeness of diurnal cycle for measuredtemperatures is estimated as the standard deviation of thedifferences between TRS and TDiur


Estimate forLauder

JordisTradowsky

Introduction

Methodology


Example ofthe SASBE

Summary

References

5/16




σfit =

√√√√ 81∑i=1

σ2ζi

(∂TDiur

∂ζi

)2

(1)




Estimate forLauder

JordisTradowsky

Introduction

Methodology


Example ofthe SASBE

Summary

References

5/16




σfit =

√√√√ 81∑i=1

σ2ζi

(∂TDiur

∂ζi

)2

(1)




Estimate forLauder

JordisTradowsky

Introduction

Methodology


Example ofthe SASBE

Summary

References

6/16

Diurnal Cycle

I Uncertainty on the diurnal cycle is estimated as:

σTDiur =√σ2

fit + σ2representativeness (2)

265

270

275

280

285

290

295

●

●

●

●●

●

●

●

●

●

●

●

●

●

●

●

●

●● ●

●

●

●

●

0 2 4 6 8 10 12 14 16 18 20 22

Tem

pera

ture

[K]

UTC hour of day

Figure 1: Diurnal cycle 925 hPa above Lauder at the 1st of January.


Estimate forLauder

JordisTradowsky

Introduction

Methodology


Example ofthe SASBE

Summary

References

7/16

Lauder component

I Lauder RS80 and RS92 radiosonde

I At the time of a measurement, SASBE temperature agreeswith measured temperature and the uncertainty agreeswith RS uncertainty

I RS80: 1-σ=0.5 K, RS92: 1-σ=0.25 K


Estimate forLauder

JordisTradowsky

Introduction

Methodology


Example ofthe SASBE

Summary

References

7/16

Lauder component



I RS80: 1-σ=0.5 K, RS92: 1-σ=0.25 K


Estimate forLauder

JordisTradowsky

Introduction

Methodology


Example ofthe SASBE

Summary

References

7/16

Lauder component



I RS80: 1-σ=0.5 K, RS92: 1-σ=0.25 K


Estimate forLauder

JordisTradowsky

Introduction

Methodology


Example ofthe SASBE

Summary

References

8/16

Invercargill component

I Temperature anomalies from Invercargill are transferred toLauder with a regression model

T ′Lau = γ︸︷︷︸Offset

+ β · T ′Inv︸︷︷︸T anomaly

+ η ·∆SP︸︷︷︸∆ surface P

+ κ ·∆ST ′︸︷︷︸∆ surface T

+ ε︸︷︷︸Residual

(3)

I Temperature anomaly term is expanded in a Fourier seriesof the wind direction


Estimate forLauder

JordisTradowsky

Introduction

Methodology


Example ofthe SASBE

Summary

References

8/16

Invercargill component

I Temperature anomalies from Invercargill are transferred toLauder with a regression model

T ′Lau = γ︸︷︷︸Offset

+ β · T ′Inv︸︷︷︸T anomaly

+ η ·∆SP︸︷︷︸∆ surface P

+ κ ·∆ST ′︸︷︷︸∆ surface T

+ ε︸︷︷︸Residual

(3)

I Temperature anomaly term is expanded in a Fourier seriesof the wind direction


Estimate forLauder

JordisTradowsky

Introduction

Methodology


Example ofthe SASBE

Summary

References

9/16

Combining the SASBE components

The SASBE combines the diurnal cycle, the Lauder componentand the Invercargill component.

TSASBE (t) = TDiur (t)︸︷︷︸Diurnal cycle

+N∑

i=1

φ · wi · T ∗iLau(t)︸︷︷︸

Lauder component

+M∑

j=1

(1 − φ) · wj · T ∗jLau(t)

︸︷︷︸Invercargill component

(4)

Using the attenuated temperature anomaly T ∗:

T ∗iLau(t) = T ′Lau(ti ) · acf (∆ti ) (5)

The uncertainties are propagated through Eq. (4).

σTSASBE (t) =

√√√√√σ2TDiur

(∂TSASBE

∂TDiur

)2

+ σ2TRSLau

(∂TSASBE

∂TRSLau

)2

+ σ2TRSLau

∂TSASBE

∂TRSLau

2

(6)


Estimate forLauder

JordisTradowsky

Introduction

Methodology


Example ofthe SASBE

Summary

References

9/16




+N∑

i=1

φ · wi · T ∗iLau(t)︸︷︷︸

Lauder component

+M∑

j=1

(1 − φ) · wj · T ∗jLau(t)


(4)




σTSASBE (t) =


(∂TSASBE

∂TDiur

)2

+ σ2TRSLau

(∂TSASBE

∂TRSLau

)2

+ σ2TRSLau

∂TSASBE

∂TRSLau

2

(6)


Estimate forLauder

JordisTradowsky

Introduction

Methodology


Example ofthe SASBE

Summary

References

9/16




+N∑

i=1

φ · wi · T ∗iLau(t)︸︷︷︸

Lauder component

+M∑

j=1

(1 − φ) · wj · T ∗jLau(t)


(4)




σTSASBE (t) =


(∂TSASBE

∂TDiur

)2

+ σ2TRSLau

(∂TSASBE

∂TRSLau

)2

+ σ2TRSLau

∂TSASBE

∂TRSLau

2

(6)


Estimate forLauder

JordisTradowsky

Introduction

Methodology


Example ofthe SASBE

Summary

References

10/16

Example of the temperature SASBE

●

●● ● ●

●

●

●

●

●

●

●

●●

●●

●

●

●●

●

●

●

●

280

285

290

295 2010−12−06

seq(0, 23)

●

●

●●

●

●

●

●

●

●

●

●

●

●

●

●

●

●

●

●

●

●

●

●

Tem

pera

ture

[K]

●

●

●

●●

●

●

●

●

●

●

●

●

●

●

●

●

●

●●

●

●

●

●

●

●

●●

●

●

●

●

●

●

●

●

●

●

●

●

●

●

●

●

●

●

●

●

● TSASBE

TDiur

TRS LauderTRS Invercargill

Tdeniala)

seq(0, 23)

01

2σ

[K]

σTSASBE

σTDiur

σLauderσInvercargill

b)−

40

4A

nom

aly

[K]

Lauder component Invercargill componentc)

0 1 2 3 4 5 6 7 8 9 10 12 14 16 18 20 22

0.5

1W

eigh

t

UTC hour of day

α 1 − α

d)


Estimate forLauder

JordisTradowsky

Introduction

Methodology


Example ofthe SASBE

Summary

References

11/16

Example of the temperature SASBE

●

●●

● ●●

●

●

●

●

●

●

●

●

●

●

●

●

● ●●

●

●

●

275

280

285

290

295 2010−12−07

seq(0, 23)

●

●

●

● ●

●

●

●

●

●

●

●

●

●

●

●

●

●

●

●

●

●

●

●

Tem

pera

ture

[K]

●

●

●

●●

●

●

●

●

●

●

●

●

●

●

●

●

●

●●

●

●

●

●

●

●

●

● ●

●

●

●

●

●

●

●

●

●

●

●

●

●

●

●

●

●

●

●

●

● TSASBE

TDiur

TRS LauderTRS Invercargill

Tdeniala)

seq(0, 23)

01

2σ

[K]

σTSASBE

σTDiur

σLauderσInvercargill

b)−

40

4A

nom

aly

[K]

Lauder component Invercargill componentc)

0 1 2 3 4 5 6 7 8 9 10 12 14 16 18 20 22

0.5

1W

eigh

t

UTC hour of day

α 1 − α

d)


Estimate forLauder

JordisTradowsky

Introduction

Methodology


Example ofthe SASBE

Summary

References

12/16

Summary

I The final version of the SASBE for Lauder has beenpublished

I The methodology has been published as discussion paper(Tradowsky et al., 2018)→ please provide feedback

I Outlook: I will keep working with Greg and colleaguesfrom around the world (GRUAN, radio occultation,measurement campaigns).→ if you are interested to cooperate on a project, pleaselet me know

I Outlook: Funding dependent: SASBEs for West AntarcticIce Sheet → radiative transfer calculation using SASBE →study the sensitivity of the radiation balance to changes inthe surface emissivity. For the motivation please seeFeldman et al. (2014).


Estimate forLauder

JordisTradowsky

Introduction

Methodology


Example ofthe SASBE

Summary

References

12/16

Summary






Estimate forLauder

JordisTradowsky

Introduction

Methodology


Example ofthe SASBE

Summary

References

12/16

Summary






Estimate forLauder

JordisTradowsky

Introduction

Methodology


Example ofthe SASBE

Summary

References

12/16

Summary






Estimate forLauder

JordisTradowsky

Introduction

Methodology


Example ofthe SASBE

Summary

References

13/16

References I

Feldman, D., Collins, W., Pincus, R., Huang, X., and Chen, X. (2014).Far-infrared surface emissivity and climate. Proc. Natl. Acad. Sci. U. S. A.,111.

Tradowsky, J., Bodeker, G., Querel, R., Builtjes, P., and Fischer, J. (2018).Combining data from the distributed gruan site lauder-invercargill, newzealand, to provide a site atmospheric state best estimate of temperature.Earth System Science Data Discussions.


Estimate forLauder

JordisTradowsky

Introduction

Methodology


Example ofthe SASBE

Summary

References

14/16

Thank you for your attention!

Documents

A Site Atmospheric State Best Estimate of Temperature for ...€¦ · A Site Atmospheric State Best Estimate of Temperature for Lauder, New Zealand Jordis Tradowsky1;2;3, Greg Bodeker1,