evaluation of the atmospheric water vapor content in the...

Evaluation of the atmospheric water vapor content in the regionalclimate model ALARO-0 using GNSS observations

Julie Berckmans1, Roeland Van Malderen1, Eric Pottiaux2, Rosa Pacione3

1 Royal Meteorological Insitute of Belgium, 2 Royal Observatory Belgium, 3 e-GEOS S.p.A. ASI/CGS Matera

17-19 May 2017

EUREF Symposium

Wroclaw

Outline

Introduction

Data

Methods

Results

Discussion and future research

J. Berckmans IWV validation Outline 2 / 17

Research topic

AimEvaluation of water vapor in regional climate models using observations

from GNSS

MotivationLack of validation by regional climate models, new reprocessed dataset

ready for climate studies

RelevanceQuality of regional climate model for climate projection

J. Berckmans IWV validation Introduction 3 / 17

Climate model

ALARO

−200

0

200

400

600

800

1000

1200

1400

1600

1800

2000

2200

2400

2600

2800

3000

3200

3400

Configuration of the ALADINmodel (v0)

Lateral boundary conditionsERA-Interim

Land surface model SURFEXDetails:Size: 149 x 149 grid points

Horizontal resolution: 20 km

Vertical 46 levels: from 17 m to 72 km

Lambert conformal projection

Radiation scheme ACRANEB

ALADIN International Team (1997), Gerard et al. (2009), DeTroch et al. (2013), Giot et al. (2016), Masson et al. (2003,Masson et al. (2013)

J. Berckmans IWV validation Data 4 / 17

GNSS Observations

EPN tropospheric product repro 2 1996-2014, selection criteria:

fit within domain

min. 10 years of data

min. 15 days per month

100 stations selected

Pacione et al. (2016)

J. Berckmans IWV validation Data 5 / 17

IWV calculation

ZTD observations to IWV

[Stull, 1995][Bevis et al., 1992]

=

[Askne and Nordius, 1987][Davis et al., 1985]

[Elgered et al., 1991]

Simplification for Tm: Hypsometric equation:

[Saastamoinen, 1972]

[Hogg et al., 1981]

Tm and Ps from ERA-Interim

J. Berckmans IWV validation Methods 6 / 17

IWV calculation

Model calculation of IWV

Horizontal interpolation: 4 nearest grid points (weighting: inversedistance)

Vertical linear interpolation based on Hagemann et al. (2003) butusing:

Pressure station level using barometric formulaT, Sfpres, H from modelStandard lapse rate for temperature -0.0065K/mVertical levels from lowest to +/- 20 km

Hagemann et al. (2003)

J. Berckmans IWV validation Methods 7 / 17

Model performance

Differences between models and observations

14 16 18 20

1416

1820

GNSS observations (kgm−²)

ER

AI /

ALA

RO

(kg

m−

2)

●

●●

●

●

●

●

●

●

●

●●

●

●

●

●

●

●●

●

●

●

●

●

●

●

●

●●

●

●

●

●

●

●

●

●

●

●

●

●

●

●

●

●

●

●

●

●

●

●

●

●

●

●

●

● ●

●

●

●

●

●

●

●

●

●

●

●

●

●

●

●

●

●

●

●

●●

●

●

●

●●

●

●

●●

●

●

●●

●

●

●

●

●

●

●●

●

●●

●

●

●

●

●

●

●

●●

●

●

●

●

●

●

●

●

●

●

● ●

●●

●

●

●

●

●

●

●

●

●

●

●

●

●

●

●●

●

●

●

●

●●

●

●

●●

●

●

●

●●

●

●

●

●●

●

● ●

●

●

●

●

●

●●

●

●

●

●

●

●

●

●

●●

●

●

●

●

●

●

●

●

●

●

●

●

●●

●

●

●●●

●

●

●

●

●

●

●

●●

●

●

●

●

●

●

●

●

●

●

●●

●

●

●●

●

●

ERAI−GNSS (0.977X+0.705, R²=0.988)

ALARO−GNSS (0.960X+0.617, R²=0.975)

●

ERAIALARO 1:1 line

−6 −4 −2 0 2 4 6 8

0.0

0.2

0.4

0.6

0.8

1.0

IWV (kgm−²)

Den

sity

ERAI−GNSS

median

−1 sigma +1 sigma

−6 −4 −2 0 2 4 6 8

0.0

0.2

0.4

0.6

0.8

1.0

IWV (kgm−²)

Den

sity

ALARO−GNSS

median

−1 sigma +1 sigma

J. Berckmans IWV validation Results 8 / 17

Model performance

Distribution of all data Distribution of the 95th percentile

0 10 20 30 40

0.00

0.02

0.04

0.06

IWV (kgm−²)

Den

sity

GNSSERAIALARO

25 30 35 40

0.00

0.05

0.10

0.15

0.20

0.25

0.30

IWV (kgm−²)

Den

sity

GNSSERAIALARO


Seasonal variability5

1015

2025

30

Month

IWV

(kg

m−

²)

J F M A M J J A S O N D

GNSSERAIALARO

● ● ● ● ● ● ● ● ● ● ● ●

−2

−1

01

23

Month

IWV

diff

eren

ce (

kgm

−²)


● ERAI−GNSSALARO−GNSS

IWV bias

Overestimation ERAI, constant

Larger standard deviation insummer, both ERAI and ALARO

ALARO performs better than ERAI,except for July-August

Large underestimation ALARO inJuly-August


Seasonal variability

●

●

● ● ●

●

●●

●

●

●●●

●

●●●●●

●

●


−50

050

100

150

200

Pre

cipi

tatio

n bi

as (

%)

●

●

●●●●

●

●

●

●

●

●

●

●

●

●●

●

●

●

●

●

●

●

●●

●

●●

●

●● ●

●●

●●●


−6

−4

−2

0

Tem

pera

ture

bia

s (°

C)

ALARO - E-OBS: precipitation bias

Good performance for May,June,Sep,Oct,Nov

Large neg. bias August

Large spread July

Large underestimation of precipitation inAugust+ Large underestimation of temperature= smaller moister holding capacity= explains negative IWV bias.


Spatial variability

ALARO - GNSSLarge outlier = SJDV

Altitudinal difference (m)

Abs

olut

e IW

V d

iffer

ence

(kg

m−

²)

0 10 20 50 100 200 500

01

23

45


Spatial variabilityBias [kgm-2] Stdev [kgm-2]

●

●

●

●

●●

●

●●●

●

●

●

●

●

●●

●

●

● ●

●

●

●

●

●●

●

●

●

●

●●

●

●

●

●

●

●

●

●

●

●●

●

●

●●

●

●

●

●

●

●

●

●

●

●●

●

●

●

●

●●

●

●

●

●

●

●

●

●

●

●

●

●

●

●

●

●

●

●

●

●●

●

●

●

●

●

●

●

●

●

●

●

●

● ●

−2

−1.5

−1

−0.5

−0.1

0.1

0.5

1

1.5

2

●

●

●

●

●●

●

●●●

●

●

●

●

●

●●

●

●

● ●

●

●

●

●

●●

●

●

●

●

●●

●

●

●

●

●

●

●

●

●

●●

●

●

●●

●

●

●

●

●

●

●

●

●

●●

●

●

●

●

●●

●

●

●

●

●

●

●

●

●

●

●

●

●

●

●

●

●

●

●

●●

●

●

●

●

●

●

●

●

●

●

●

●

● ●

0.5

1

1.5

2

2.5

3

3.5

4

4.5

5

5.5

●

●

●

●

●●

●

●●●

●

●

●

●

●

●●

●

●

● ●

●

●

●

●

●●

●

●

●

●

●●

●

●

●

●

●

●

●

●

●

●●

●

●

●●

●

●

●

●

●

●

●

●

●

●●

●

●

●

●

●●

●

●

●

●

●

●

●

●

●

●

●

●

●

●

●

●

●

●

●

●●

●

●

●

●

●

●

●

●

●

●

●

●

● ●

●

●

●

●

●●

●

●●●

●

●

●

●

●

●●

●

●

● ●

●

●

●

●

●●

●

●

●

●

●●

●

●

●

●

●

●

●

●

●

●●

●

●

●●

●

●

●

●

●

●

●

●

●

●●

●

●

●

●

●●

●

●

●

●

●

●

●

●

●

●

●

●

●

●

●

●

●

●

●

●●

●

●

●

●

●

●

●

●

●

●

●

●

● ●

ERAI ERAI

ALARO ALARO


Discussion

Overestimation ERAI ≈ Lucas-Picher et al. (2013)

Larger standard deviation is expected with regional model comparedwith ERAI

Larger standard deviations in summer for both ALARO and ERAI

Underestimation of regional climate model in summer

Similar results as in ≈ Ning et al. (2013), but based on differentGNSS dataset and regional climate model

Relation precipitation and temperature model bias with IWV bias

Largest differences ALARO and ERAI in southern Europe = drymodel bias

Latitudinal dependence ≈ Pacione et al. (2016)

J. Berckmans IWV validation Conclusions 14 / 17

Future research

Closer look at spatial variability

Closer look at intra-month variability

Group stations based on similar characteristics

Diurnal cycle

...


References

Hagemann, S., Chen, C., Clark, D. B., Folwell, S., Gosling, S. N.,Haddeland, I., Hanasaki, N., Heinke, J., Ludwig, F., Voss, F., andWiltshire, A. J.: Climate change impact on available water resourcesobtained using multiple global climate and hydrology models, Earth Syst.Dynam., 4, 129-144, doi:10.5194/esd-4-129-2013, 2013.

Ning, T., Elgered, G., Willen, U., Johansson, J.M.: Evaluation ofthe atmospheric water vapor content in a regional climate model usingground-based GPS measurements, J. Geophys. Res., 118, 329-339, doi:10.1029/2012DJ018053, 2013.

Pacione, R., Araszkiewicz, A., Brockmann, E., and Dousa, J.:EPN-Repro2: A reference GNSS tropospheric data set over Europe, Atmos.Meas. Tech., 10, 1689-1705, doi:10.5194/amt-10-1689-2017, 2017.


Extra: precipitation bias summer

−100

−90

−80

−70

−60

−50

−40

−30

−20

−10

0

10

20

30

40

50

60

70

80

90

100


evaluation of the atmospheric water vapor content in the...

Documents