La lecture à portée de main
Découvre YouScribe en t'inscrivant gratuitement
Je m'inscrisDécouvre YouScribe en t'inscrivant gratuitement
Je m'inscrisDescription
Sujets
Informations
Publié par | universitat_bremen |
Publié le | 01 janvier 2010 |
Nombre de lectures | 23 |
Langue | English |
Poids de l'ouvrage | 30 Mo |
Extrait
Sea Level Variations
derived from Mass Conserving
Finite Element Sea-Ice Ocean Model
Study of Major Contributions to Sea Level Change
in the Recent Past
Sandra-Esther Brunnabend
Universit¨at Bremen 2010Sea Level Variations
derived from Mass Conserving
Finite Element Sea-Ice Ocean Model
Study of Major Contributions to Sea Level Change
in the Recent Past
Vom Fachbereich fur¨ Physik und Elektrotechnik
der Universit¨ at Bremen
zur Erlangung des akademischen Grades eines
Doktor der Naturwissenschaften (Dr. rer. nat.)
genehmigte Dissertation
von
Sandra-Esther Brunnabend M.Sc. (ESPACE)
aus Bremerhaven
1. Gutachter: Prof. Dr. Peter Lemke
2.hter: Dr.-Ing. Jurgen¨ Kusche
Eingereicht am: 02. Dezember 2010
Tag des Promotionskolloquiums: 17. Februar 2011Abstract
During the last century sea level rise strongly increased compared to sea level change in
the last 2000 years. The present study investigates global and regional sea level change,
simulated with the finite element sea-ice ocean model (FESOM). The major goal is to
separate sea level change into steric and eustatic contributions and to estimate the influence
of Greenland and Antarctic ice sheet melt on global and regional sea level.
Modeled steric height variations show realistic regional geophysical patterns compared
with steric height variations derived from altimetry measurements and GRACE. Compared
to the time before the 1990’s, an increased global trend in steric sea level rise is found
in estimates derived from the model and from satellite measurements. Modeled ocean
mass exhibits reasonable spatial structures. However, the trend in the global model mean
cannot be trusted in FESOM as it strongly depends on the mass budget of the model,
which is determined by uncertain mass fluxes. To account for this, global mean ocean
mass variations need to be optimized to realistic values. To this end results from GRACE
in combination with GPS data is used.
Greenland and Antarctic ice sheet melting influence the global sea level mainly through
the additional mass. The eustatic sea level rises by about 0.3 mm/yr for 100 Gt/yr of
melt water. Additionally, the fresh water causes local steric variations in sea level that
are transported farther by ocean currents. The ice sheet mass loss yields a decrease in
gravitational attraction causing a sea level fall near the source of mass loss but also to a
slight increase at long distance. This effect is computed for the Greenland ice sheet mass
loss using Green’s functions. It leads to a decreased sea level near the Greenland coast
and to a slightly increased sea level in the Southern Ocean. The effect of different melting
scenarios is investigated.
iZusammenfassung
W¨ ahrend des letzten Jahrhunderts ist der Meeresspiegel st¨arker angestiegen als in den ver-
gangenen 2000 Jahren. Diese Studie untersucht globale und regionale Veranderungen¨ des
Meeresspiegels, die mit dem Finite-Elemente-Meereis-Ozean-Modell (FESOM) simuliert
werden. Das Hauptziel ist es, sterische und eustatische Beitr¨ age in den Meerespiegel-
¨anderungen zu separieren. Ausserdem wird der Einfluss des Schmelzwassers der beiden
grossen Eisschilde auf Gr¨onland und der Antarktis auf den regionalen Meeresspiegel un-
tersucht.
Modellierte sterische H¨ohenanderungen¨ zeigen realistische regionale geophysikalische
Strukturen, a¨hnlich denen, die aus Altimetriedaten und GRACE abgeleitet worden sind.
Verglichen mit der Zeit vor den 1990er Jahren, ist ein erh¨ ohter Trend im globalen sterischen
Meeresspiegelanstieg sowohl in den Modellergebnissen als auch in den Satellitenmessungen
zu beobachten. Modellierte Ozeanmassenvariationen zeigen angemessene r¨aumliche Struk-
turen. Dem Trend der globalem mittleren Massenvariationen ist normalerweise mit einem
grossen Fehler versehen, da dieser stark von der Massenbilanz des Modells abh¨ angt, die
durch unsichere Massenflusse¨ bestimmt ist. Daher muss die Beschreibung der globalen
mittleren Massenvariationen optimiert werden. Hierfur¨ werden Ergebnisse von GRACE in
Kombination mit GPS Daten verwendet.
Das Schmelzwasser der Eisschilde in Gr¨onland und der Antarktis beeinflusst den glob-
alen Meeresspiegel. Die Simulationen zeigen einen eustatischen Meeresspiegelanstieg von
etwa 0,3 mm pro Jahr, wenn Schmelzwasser in Hohe¨ von 100 Gt pro Jahr in den Ozean
fliessen. Das zus¨ atzliche Sussw¨ asser verursacht lokale sterische Variationen des Meer-
esspiegels, die durch die Meeresstr¨ omungen in weiter entfernte Regionen transportiert wer-
den. Ausserdem verursacht der Massenverlust eine Verringerung der Anziehungskraft des
Eisschildes. Dies fuhrt¨ zu einem verringerten Meeresspiegel in der N¨ ahe des Eisschildes
und zu einer leichten Erh¨ ohung in weiter entfernten Regionen. Dieser Effekt wird fur¨ die
Gr¨ onlandeisschmelze mit Hilfe von Green’s Funktionen berechnet. Er verursacht einen
Meeresspiegelfall in der N¨ ahe der Gr¨ onl¨ andischen Kuste¨ und einem leichten Anstieg im
iiiii
S¨ udozean. Der Effekt von verschiedenen Schmelzszenarien wird untersucht.Contents
1 Introduction 1
2ModelDescription 6
2.1 Method of Finite Elements ........................... 6
2.2 Finite Element Sea-Ice Ocean Model . 7
2.2.1 Finite Element Sea-Ice Model . 7
2.2.2 Finite Element Ocean Model...................... 8
2.2.3 Coupling............. 10
2.3 Subgrid-Scale Processes ........ 12
2.4 Surface Heat Budget .............................. 13
2.5 Sea Level Equation ........... 14
2.6 Spatial Discretization 15
2.7 Atmospheric Forcing .............................. 16
2.7.1 NCEP/NCAR Data ...... 18
2.7.2 ERA40 Data 18
2.7.3 Operational ECMWF data....................... 19
2.7.4 ERA Interim .......... 20
2.7.5 GPCP Precipitation ...... 21
2.7.6 Reference Model Simulation and Sensitivity Experiments ...... 22
3 Variations of Ocean Mass 23
3.1 Land-Ocean Mass Exchange .......................... 24
3.2 Mass Exchange between Atmosphere and Ocean ...... 27
3.3 Mass Balance in FESOM........ 30
3.4 Weekly Ocean Bottom Pressure Anomalies .................. 3
3.5 Error Estimation of Modeled Ocean Mass Variations ... 35
3.5.1 Influence of Spatial Discretization .......... 36
ivCONTENTS v
3.5.2 FESOM and the Large Scale Geostrophic Model........... 38
3.5.3 Influence of Atmospheric Forcing .......... 40
3.6 Gravity Recovery and Climate Experiment ......... 42
3.7 Joint Inversion ................................. 48
3.7.1 Impacts of modeled OBP Error to the Inversion .. 49
3.7.2 Global Mass Correction derived from Inversion... 50
3.8 Ocean Bottom Pressure Recorders....................... 53
3.8.1 Correlation with OBPR .... 54
3.8.2 OBPR for Regional Mean Mass Variations ..... 60
3.8.3 Validation of modeled OBP error ................... 6
3.9 Conclusion........................... 68
4 Sea Surface Topography and Steric Sea Level 70
4.1 Modeled Surface Heat Flux ............... 71
4.2 Surface Heat Flux Correction ................ 75
4.3 Modeled Sea Surface Topography ... 76
4.4 Variations in Sea Level .................. 78
4.5 Modeled Steric Height Variations .............. 81
4.6 Satellite Altimetry ........... 83
4.7 Multi-Mission Altimetry Measurements......... 85
4.8 Identification of Modeled Steric Height Changes in Altimetry........ 86
4.9 Conclusion........................... 8
5 Sea Level Change due to Ice Sheet Melting 90
5.1 Definition of Ice sheets .................. 91
5.2 Loading and Self Attraction ................. 92
5.3 Melting of the Antarctic Ice sheet ... 94
5.4 Mass Variations of the Greenland Ice Sheet.................. 97
5.4.1 Constant Melt Rates ................. 9
5.4.2 Sea Level Evolution in Coastal Regions .......104
5.4.3 Varying Melt Rates ................105
5.5 Conclusion...........................108
6 Summary and Outlook 111
Bibliography 118CONTENTS vi
Appendix 133
AListofAcronyms................................13
B Global Positioning System .......135
C Comparison of OBPR and FESOM Timeseries .......136
D C of time series from OBPR and Inverse Solutions ........149
E Comparison of time series from OBPR and GRACE S .......162
F Comparison of time series from OBPR and corrected FESOM175List of Figures
1.1 Satellite measurements and ocean modelling ................. 5
2.1 Spatial discretization and bottom topography ....... 15
2.2 Accumulation of water fluxes.......................... 20
2.3 Global Sum of GPCP Precipitation .. 21
3.1 River runoff ........................ 25
3.2 Global river runoff........... 26
3.3 Difference in SSH due to different river runoff ....... 27
3.4 Global integral of precipitation minus evaporation .............. 28
3.5 Ten year mean precipitation ................. 29
3.6 Correlation of 10yr mean precipitation to GPCP...... 30
3.7 Ratio of 10yr mean precipitation from GPCP and ECMWF ........ 30
3.8 Global fresh water flux ............................. 31
3.9 Variations of global mean ocean mass . 32
3.10 Variations in ocean bottom pressure, if the heat flux is set to 0....... 33
3.11 Global mean OBP variations modeled with FESOM ............. 34
3.12 Weekly mean pressure of ocean and atmosphere ...... 35
3.1