Flechtner F.M., Gruber Th., G?ntner A., Mandea M., Rothacher M., Sch?ne T., Wickert J. (Eds.) System Earth via Geodetic-Geophysical Space Techniques

Подождите немного. Документ загружается.

Using ARGO, GRACE and Altimetry Data t o Assess the Quasi Stationary 355

Fig. 4 Inter-annual change in

altimetry data (2006–2005)

Fig. 5 Change in barotropic stream function (ARGO06-ARGO05 left and AltARGO06-

AltARGO05 right). Altimetry has a stronger impact on total transports then ARGO data have

more pronounced changes in AltARGO experiments. The changes in T and S ﬁelds

are distinct and quite similar between ARGO only experiments and AltARGO solu-

tions. In accordance to these ﬁndings, only smaller modiﬁcations in T and S are

needed to compensate for the misﬁt in surface elevation. The variability in transports

can be either explained by the bottom pressure change (which is loosely constrained

during the assimilation to that of a forward solution; Sidorenko et al., 2005) or the

change in vertically integrated potential energy (which is the result of redistribution

of T and S within the water column). Figure 6 shows the difference in vertically

integrated potential energy between ARGO06-ARGO05 (left panel) and between

AltARGO06-AltARGO05 (right panel).

It demonstrates that a signiﬁcant part of the transport variability is due to the

variability in vertically integrated potential energy. Thus, despite the fact that T

and S adjustments done by IFEOM are mainly due to the ARGO anomalies, the

modiﬁcations in potential energy are mainly from including altimetry. The latter in

turn has a signiﬁcant impact on transport variability.

Figure 7 shows the temperature change (AltARGO06-AltARGO05) across a sec-

tion at 37.5

◦

N. It is dominated by the ARGO anomalies. The changes due to the

altimetry assimilation are an order of magnitude smaller and are of the same sign

356 F. Richter et al.

Fig. 6 Change in vertical integrated potential energy (ARGO06-ARGO05 left and AltARGO06-

AltARGO05 right) scaled by 10

–5

. The transport changes induced by using altimeter data are

produced by adjusting the vertical density structure in the model

Fig. 7 Temperature change along a section through the North Atlantic at 37.5

◦

N. Difference

between solutions AltARGO06-AltARGO05 is shown in background colour. The difference in

the observations (2006–2005) at ARGO positions is depicted by colour coding the following sym-

bols: both years are present in ARGO data (circle), if only 1 year (2005 square, 2006 triangle)is

present the difference is computed using the climatology data instead of the missing year

within the entire depth (not shown here). The plot demonstrates that the model takes

into account even the transient signal which can be present in raw data. The pattern

of positive anomaly at about 40

◦

W corresponds to an unstable proﬁle measurement

(year 2005). This anomaly is partly captured by IFEOM (which accounts for the

whole available information) and results in stable model solution.

4 Summary and Conclusions

The quasi-stationary circulation of the North Atlantic is estimated for the years

2005–2006 using ARGO, GRACE and altimetry data. The assimilation technique

has been used to reduce the misﬁt between data and model by adjusting T and S

Using ARGO, GRACE and Altimetry Data t o Assess the Quasi Stationary 357

ﬁelds. It is shown that inclusion of the additional information from altimetry sig-

niﬁcantly changes the estimates of transports and variability. The sea level rise

between 2005 and 2006 of about 5 cm in the Nordic seas is shown to be present

in both, ARGO and altimetry data. Both datasets also reveal large variability in

the Gulf Stream region and its extension which contributes to the estimated inter-

annual differences of about 20 cm. The Gulf Stream region shows the transport

variability of about 20 Sv in maximum. The changes are more pronounced for the

simulations including altimetry data despite the fact that using altimetry leads to

only small additional adjustments in T and S. These adjustments in turn modify

the vertically integrated potential energy signiﬁcantly, which explains the transport

variability. Thus, the vertically integrated potential energy anomaly shows that the

transport variability is dominated mainly by the altimetry information while the

major changes in T and S are explained by the ARGO data.

Acknowledgment This is publication no. GEOTECH 1249 of the GEOTECHNO-LOGIEN

programme of BMBF, grant 03F0434B.

References

Ganachaud A, Wunsch C (2003) Large-scale ocean heat and freshwater transports during the world

ocean circulation experiment. J. Clim. 16, 696–705.

Gouretski VV, Koltermann KP (2004) WOCE Global Hydrographic Climatology. Bundesamt für

Seeschifffahrt und Hydrographie, Hamburg und Rostock, Germany.

Guinehut S, Coatanoan C, Dhomps A-L, Le Traon P-Y, Larnicol G (2009) On the use of satellite

altimeter data in ARGO quality control. J. Atmos. Oceanic Technol. 26, 395–402.

Ivchenko VO, Danilov SD, Sidorenko DV, Schröter J, Wenzel M, Aleynik DL (2007)

Comparing the steric height in the Northern Atlantic with satellite altimetry. Ocean Sci. 3,

485–490.

Kivman G, Danilov S, Fritzsch B, Harig S, Reick C, Schröter J, Seufer V, Sidorenko D, Staneva

J (2005) Improved estimates of the oceanic circulation using the CHAMP geoid. Earth

Observation with CHAMP, Results from Three Years in Orbit, Springer-Verlag, Heidelberg,

pp. 211–216.

Losch M, Schröter J (2004) Estimating the circulation from hydrography and satellite altimetry in

the southern ocean: limitations imposed by the current geoid models. Deep-Sea Res. I 51(9),

1131–1143.

Lumpkin R, Speer K (2007) Global ocean meridional overturning. J. Phys. Oceanogr. 37,

2550–2562.

Reigber C, Schmidt R, Flechtner F, König R, Meyer U, Neumayer K-H, Schwintzer P, Yuan Zhu S

(2005) An earth gravity ﬁeld model complete to degree and order 150 from GRACE: EIGEN-

GRACE02S. J. Geodyn. 39(1), 1–10.

Rio MH, Hernandez F (2004) A mean dynamic topography computed over the world ocean from

altimetry, in situ measurements, and a geoid model. J. Geophys. Res. 109, C12032.

Rio MH, Schaeffer P, Lemoine JM, Hernandez F (2005) Estimation of the ocean Mean Dynamic

Topography through the combination of altimetric data, in-situ measurements and GRACE

geoid: from global to regional studies. Proceedings of the GOCINA international workshop,

Luxembourg.

Sidorenko D, Danilov S, Kivman G, Schröter J (2005) On the use of a deep pressure gradient

constraint for estimating the steady state ocean circulation from hydrographic data. Geophys.

Res. Lett., doi: 10.1029/2005GL024716.

358 F. Richter et al.

Stammer D, Wunsch C, Giering C, Eckart C, Heimbach P, Marotzke J, Adcroft A, Hill CN,

Marshall J (2002) The global ocean circulation during 1992–1997, estimated from ocean

observations and a general circulation model. J. Geophys. Res. 107(C9), 3118–3145.

Stammer D, Wunsch C, Giering C, Eckart C, Heimbach P, Marotzke J, Adcroft A, Hill CN,

Marshall J (2003) Volume, heat and freshwater transports of the global ocean circulation 1993–

2000, estimated from a general circulation model constrained by World Ocean Circulation

Experiment (WOCE) data. J. Geophys. Res. C1, 3007.

Treguier AM, Theetten S, Chassignet EP, Penduff T, Smith R, Talley L, Beismann JO, Böning C

(2005) The north Atlantic subpolar gyre in four high-resolution models. J. Phys. Oceanogr. 35,

757–774.

Wenzel M, Schröter J (1998) Combined assimilation of altimetric and hydrographic data. AVISO

Newsletter 6, 122–124.

Wenzel M, Schröter J, Olbers D (2001) The global ocean circulation during 1992–1997, estimated

from ocean observations and a general circulation model. Prog. Oceanogr. 48(C9), 73–119.

Wenzel M, Schröter J (2002) Assimilation of TOPEX/Poseidon data in a global ocean model:

differences in 1995–1996. Phys. Chem. Earth 27, 1433–1437.

A 15-Year Reconstruction of Sea Level

Anomalies Using Radar Altimetry

and GPS-Corrected Tide Gauge Data

Nana Schön, Saskia Esselborn, and Tilo Schöne

1 Introduction

When estimating sea level changes over the last century, one major problem is the

spatial non-uniformity. Regional estimates give a broad range of results for sea level

rise. For the determination of consistent time series, satellite radar altimetry data

(RA) with its short but near-global coverage must be combined with the relatively

long but spatially sparse set of tide gauge time series. Church et al. (2004) devel-

oped a reconstruction scheme that takes i nto account the regional differences and

can produce reliable estimates. In this paper, a similar reconstruction procedure is

used, but improved in two aspects. First, the typical correction for land movement,

commonly accomplished by employing a global isostatic adjustment (GIA) model

(e.g., Davis and Mitrovica, 1996; Lambeck et al., 1998; Peltier, 2004), is replaced by

applying vertical land movement trends extracted from the combined TIGA (GPS

Tide Gauge Benchmark Monitoring – Pilot Project) dataset of co-located GPS sta-

tions to the tide gauge data set (Schöne et al., 2009). Second, a careful selection of

the tide gauges used in the reconstruction minimizes the inﬂuence of their spatial

distribution.

2 Methodology

For the reconstruction of sea level anomalies (SLA), a modiﬁed Optimal

Interpolation (OI) algorithm (Kaplan et al., 1997) is applied. Fully corrected and

harmonized monthly altimetric sea surface height anomaly data derived from the

Altimeter Database and Processing System ADS (chapter “Radar Altimetry Derived

Sea Level Anomalies – The Beneﬁt of New Orbits and Harmonization” by Schöne

et al., this issue) are gridded into 1

◦

× 1

◦

maps and are used t o estimate the global

N. Schön (B)

Helmholtz Centre Potsdam, GFZ German Research Centre for Geosciences,

Department 1: Geodesy and Remote Sensing, Telegrafenberg, 14473 Potsdam, Germany

e-mail: nana.schoen@gfz-potsdam.de

359

F. Flechtner et al. (eds.), System Earth via Geodetic-Geophysical Space Techniques,

Advanced Technologies in Earth Sciences, DOI 10.1007/978-3-642-10228-8_30,



Springer-Verlag Berlin Heidelberg 2010

360 N. Schön et al.

covariance structure as expressed in empirical orthogonal functions (EOFs). The

amplitude of these EOFs is estimated using tide gauge records. This way, the main

spatial patterns of the SLAs are extracted from the RA data, while their amplitude

time series is reconstructed using the tide gauge data. For this reconstruction, the

ﬁrst 25 area-weighted EOFs are used, which cover over 91% of the global variabil-

ity. Church et al. (2004) have argued that this number of EOFs is sufﬁcient for an

OI reconstruction.

The land movement part of the tide gauge records is GPS-corrected using verti-

cal trends derived from the TIGA combination (chapter “Reanalysis of GPS Data at

Tide Gauges and the Combination for the IGS TIGA Pilot Project” by Rudenko et al.

this issue). For this study, 18 tide gauges (see Schöne et al., 2009) are selected

from the global TIGA network. Tide gauge data for the period 1987–2001 are taken

from the PSMSL data archive (Woodworth and Player, 2003). The selection cri-

teria are chosen such as to comply with the recommendations of Douglas (2001)

for tide gauges, as well as Blewitt and Lavallée (2002) for the GPS time series.

An additional three tide gauges (La Libertad 2 (PSMSL ID 845/012), Pt. La Rue

(PSMSL ID 442/007), Unalaska (PSMSL ID 820/021)) were used for the stabiliza-

tion of the OI algorithm (see Schöne et al., 2009 for a detailed discussion). Only a

small number of tide gauges satisfy all the necessary selection criteria. In addition

to this, the inﬂuence of their geographical distribution on the reconstruction hinders

a more ﬂexible selection of tide gauges. Adaptively using all tide gauges available

for a certain period of time is not recommended, since the i ntegrity of a changing

geographical distribution in time cannot be determined. Instead, a ﬁxed conﬁgura-

tion of tide gauges that must stay unchanged throughout the reconstruction period

is chosen beforehand, which of course further diminishes the set of eligible tide

gauges.

For all sea level time series only one average GPS value for the vertical

velocity is applied. A comparison of the GPS land movement trends with those

determined by Peltier (2004) in his VM4 GIA model (Fig. 1), shows that, espe-

cially in the Southern Hemisphere, the model trends tend to be smaller than

those derived from GPS data. However, since several GPS-derived trends are

extracted from very short time series only, a fully reliable comparison cannot yet be

drawn.

It is assumed that the GPS vertical trends describe ongoing, long-term processes.

Hence, the corrections for land movement are applied not only for the period of

time over which the trend was determined, but are also projected into the past. The

respective corrections are applied from the beginning of the reconstruction period



Fig. 1 (a) Sea level anomalies (SLA) for January 1998, TOPEX radar altimetry data. (b)Sealevel

anomalies (SLA) reconstruction for January 1998 using 18 (21) tide gauges with GPS vertical

corrections. Tide gauges used in the reconstruction are depicted as stars.(c) Sea level anomalies

(SLA) reconstruction for January 1998 using 18 (21) tide gauges with GIA model vertical correc-

tions (Peltier, 2004). Tide gauges used in the reconstruction are depicted as stars. Three tide gauges

without GPS or GIA corrections were used for stabilization of the algorithm

A 15-Year Reconstruction of Sea Level Anomalies 361

Fig. 1 (Contd.)

362 N. Schön et al.

(1987) to its end in 2001. An exception is Pohnpei, where a strong positive trend is

derived from the GPS time series. Careful examination of the PSMSL time series

suggests that the strong uplift signal is the result of a recent process and that the

GPS correction should be applied only from 1999 onwards.

3 Comparison of GIA vs. GPS Corrections

The reconstruction results using GPS vertical corrections (Fig. 1b) are compared

with the one using GIA model corrections (Fig. 1c) as well as the original TOPEX

altimetry data (Fig. 1a) for January 1998. Here, t he two dominant signals are the

double tongue-shaped positive anomaly resulting from the 1997/1998 El Niño event,

and the Indian Ocean Dipole event (IOD) (Feng and Meyers, 2003) with its char-

acteristic warming (positive anomaly) in the Indian Ocean, as well as a distinct

cooling (negative anomaly) in the Indonesian/Western Paciﬁc region. Both recon-

structions agree well with the original data, with RMS differences of 3.74 cm (GPS

corrections) and 4.01 cm (GIA corrections), respectively.

In the comparison of the reconstructions, the advantage of the GPS-corrected

reconstruction is most visible in the rendition of the negative anomaly east of

Japan. Also, the intensity of the double tongue (El Niño signal) is reproduced more

distinctly. In contrast, the GIA-corrected reconstruction overestimates the positive

anomalies in the North Atlantic and also shows slightly more negative artifacts

around New Zealand. Figure 2 shows a comparison of monthly RMS values for

both reconstructions. Although, in the global scale (Fig. 2a), the RMS for the GPS

corrections is slightly higher, the regional RMS comparisons for the North Atlantic

(Fig. 2b) and North Paciﬁc (Fig. 2c) show the overall improvement from the GPS

correction.

In the Southern Paciﬁc region, the GIA-corrected gives a better result. This may

be due to the fact that some of the GPS trends used were derived from short time

series (see Schöne et al., 2009). A recalculation has now produced signiﬁcantly

lower trends for the short time series (D. Thaller, personal communication May,

2009). Hence, the results for the Southern Paciﬁc, as well as the Southern Atlantic,

should be treated with care.

The effect of the GPS vs. GIA corrections on the sea level trend is depicted

in Figure 3. A comparison with the trend extracted from the satellite altimetry data

over the common period (1995–2001) gives a very good agreement (2.2 mm/year vs.

2.4 mm/year); with the GPS-corrected reconstruction slightly closer to the altimetry

data over the whole period. The 21 uncorrected tide gauges used in the reconstruc-

tion give a considerably higher trend of 3.1 cm/year for the same period. The fact

that the rendition of the seasonal cycle is not yet satisfactory is most probably due

to the small number of tide gauges used. The lack of gauges causes an erroneous

overestimation of anomalies, resulting in a noisy reconstruction (see also Chambers

et al., 2002). The trend for the whole 15-year reconstruction period (1987–2001), is

comparably low at only 0.4 mm/year for both corrections, compared to 3.6 mm/year

A 15-Year Reconstruction of Sea Level Anomalies 363

Fig. 2 Monthly RMS time series for TOPEX/Poseidon – GPS/GIA, resp. (a) global (b)North

Atlantic (50

◦

–20

◦

Lat, 280

◦

–360

◦

Lon) (c) North Paciﬁc (50

◦

–30

◦

Lat, 150

◦

–210

◦

Lon) (d)RMS

difference GIA-GPS. The jump in 1999 marks the application of the Pohnpei GPS correction

from the tide gauges alone. It remains to be seen whether this low rate holds when

more land movement data becomes available, or whether it is, at least in part, caused

by an instability of the algorithm.

4 Conclusion

GPS corrections can markedly improve the quality of an SLA reconstruction pro-

vided that enough tide gauges are available. Especially in the North Atlantic and

North Paciﬁc, a regional improvement can be achieved (see also Snay et al., 2007).

Results for the Southern Oceans must be treated with care due to the insufﬁcient

length of GPS time series in this area (see also Blewitt and Lavallée, 2002).

Acknowledgments The author would like to thank Daniela Thaller for providing the GPS data,

as well as for her overall help and expertise. This is publication no. GEOTECH-1251 of the

GEOTECHNOLOGIEN programme of BMBF, grant 03F0434A.

364 N. Schön et al.

Fig. 3 (a) A comparison of global SLA trends for the 1995–2001 period. (b) Global SLA trends

for the 1987–2001 period

References

Blewitt G, Lavallée D (2002) Effect of annual signals on geodetic velocity. J. Geophys. Res.

107(B7), doi: 10.1029/2001JB000570.

Chambers DP, Melhaff CA, Urban TJ, Fuji D, Nerem RS (2002) Low-frequency variations in

global mean sea level: 1950–2000. J. Geophys. Res. 107(C4), 3026–3035.

Church A, White NJ, Coleman R, Lambeck K, Mitrovica JX (2004) Estimates of the regional

distribution of sea level rise over the 1950–2000 period. J. Clim. 17, 2609–2625.

Davis JL, Mitrovica JX (1996) Glacial isostatic adjustment and the anomalous tide gauge record

of eastern North America. Nature 379, 331–333.

Douglas BC (2001) Sea level change in the era of the recording tide gauge. In: Douglas BC,

Kearney MS, Leatherman SP (eds.), Sea Level Rise – History and Consequences, Academic

Press, New York, pp. 37–64.

Feng M, Meyers G (2003) Interannual variability in the tropical Indian Ocean: a two-year time-

scale of Indian Ocean Dipole. Deep-Sea Res. II 50, 2263–2284.

Kaplan A, Kushnir Y, Cane MA, Blumenthal MB (1997) Reduced space optimal analysis for

historical data sets: 136 years of Atlantic sea surface temperatures. J. Geophys. Res. 102,

27835–27860.

Lambeck K, Johnston P, Smither C (1998) Sea-level change, glacial rebound and mantle viscosity

for northern Europe. Geophys. J. Int. 134, 102–144.

Peltier WR (2004) Global glacial isostasy and the surface of the ice-age earth: the ICE-5G (VM2)

Model and GRACE, Ann. Rev. Earth Planet. Sci. 32, 111–149.

Schöne T, Schön N, Thaller D (2009) IGS tide gauge benchmark monitoring pilot project (TIGA):

scientiﬁc beneﬁts. J. Geod. 83(3–4), March 2008, doi: 10.1007/s00190-008-0269-y.

Snay R, Cline M, Dillinger W, Foote R, Hilla S, Kass W, Ray J, Rohde J, Sella G, Soler T (2007)