*Manuscript Click here to view linked References

1

The Lofoten Vortex of the Nordic Seas

2

Roshin P. Raj a, b, c,*, Léon Chafik d, J. Even Ø. Nilsen a, c, Tor Eldevik b, c, Issufo Halo e,f

3

a

Nansen Environmental and Remote Sensing Center, Thormøhlens gate 47, Bergen, Norway

4

b

Geophysical Institute, University of Bergen, Allégaten 70, Bergen, Norway

5

c

Bjerknes Center for Climate Research, Allégaten 70, Bergen, 5007, Norway

6

d

Department of Meteorology, Stockholm University, S-106 91, Stockholm, Sweden

7

e

Department of Oceanography, University of Cape Town, 7701, Rondebosch, South Africa

8

f

Nansen-Tutu Centre for Marine Environmental Research, 7701, Rondebosch, South Africa

9 10

* Corresponding author

11

Nansen Environmental and Remote Sensing Center, Thormøhlens gate 47, Bergen,

12

Norway

13

Email: [email protected]

14

Phone: +4745263862

15 16 17 18 19 20 21 22 23

Atlantic Water (AW), Eddy kinetic energy (EKE), Eddy intensity (EI), Absolute dynamic topography (ADT)

24

Abstract

25

The Lofoten Basin is the largest reservoir of ocean heat in the Nordic Seas. A particular feature

26

of the basin is ‘the Lofoten Vortex’, a most anomalous mesoscale structure in the Nordic Seas.

27

The vortex resides in one of the major winter convection sites in the Norwegian Sea, and is likely

28

to influence the dense water formation of the region. Here, we document this quasi-permanent

29

anticyclonic vortex using hydrographic and satellite observations. The vortex’ uniqueness in the

30

Nordic Seas, its surface characteristics on seasonal, inter-annual, and climatological time-scales,

31

are examined together with the main forcing mechanisms acting on it. The influence of the

32

vortex on the hydrography of the Lofoten Basin is also shown. We show that the Atlantic Water

33

in the Nordic Seas penetrate the deepest inside the Lofoten Vortex, and confirm the persistent

34

existence of the vortex in the deepest part of the Lofoten Basin, its dominant cyclonic drift and

35

reveal seasonality in its eddy intensity with maximum during late winter and minimum during

36

late autumn. Eddy merging processes are studied in detail, and mergers by eddies from the slope

37

current are found to provide anticyclonic vorticity to the Lofoten Vortex.

38

Keywords: Norwegian Atlantic Current, Lofoten Vortex, eddy kinetic energy, convection,

39

Lofoten Basin, heat loss.

40 41 42 43 44 45 2

46

1. Introduction

47

The warm and saline Atlantic Water (AW) entering the Norwegian Sea transports heat

48

towards the Arctic, and is thus a key component in maintaining the region’s relatively mild

49

climate and ice-free oceans (Rhines et al., 2008). Heat loss to the atmosphere is one of the major

50

processes resulting in water mass transformation in the Nordic Seas, an important component of

51

the Atlantic Meridional Overturning Circulation (e.g., Eldevik and Nilsen, 2013). The Lofoten

52

Basin of the Norwegian Sea (Fig. 1) is both the largest heat reservoir of the Nordic Seas

53

(Blindheim and Rey, 2004) and a region with strong atmospheric heat loss (Bunker, 1976). The

54

Lofoten Basin has earlier been identified as the region of highest eddy activity in the Nordic Seas

55

(Poulain et al., 1996). The residence time of AW in the Lofoten Basin resulting in storage of

56

large quantities of AW, is longer than any other region in the Nordic Seas, due to the deep

57

cyclonic recirculation prevailing there (Orvik, 2004; Gascard and Mork, 2008). The long

58

residence time of AW in the Lofoten Basin results in additional cooling of AW before it reaches

59

the Arctic proper. However, the details of circulation and water mass transformation in the

60

Lofoten Basin are still to a large extent unknown.

61

The Lofoten Basin, bordered by the baroclinic Norwegian Atlantic Front Current (the front

62

current) on the western side and barotropic Norwegian Atlantic Slope Current (the slope current)

63

on the eastern side (Orvik and Niiler, 2002; Mork and Skagseth, 2010), seats a large anticyclonic

64

vortex in its western part. This vortex have been reported in several studies (Ivanov and

65

Korablev, 1995a, b; Köhl, 2007; Rossby et al., 2009; Voet et al., 2010; Andersson et al., 2011;

66

Koszalka et al., 2011; Volkov et al., 2013; Søiland and Rossby, 2013). Ivanov and Korablev

67

(1995a, b) first identified this feature and reported it as a quasi-permanent anticyclonic vortex.

68

They concluded that the main forcing mechanism responsible for the stability of the vortex is 3

69

winter convection, which via deepening of isopycnals regenerates the lens and reinforces the

70

vortex’s circulation. Köhl (2007), using a numerical ocean model studied the generation

71

mechanisms and conditions for stability of the vortex. He proposed that the vortex feed energy

72

primarily from anticyclonic eddies shed from the slope current, propagating south-westward into

73

the central Lofoten Basin. Rossby et al. (2009) supported this argument by showing a link in the

74

isopycnals between the Lofoten Escarpment and deep Lofoten Basin. A recent study of the

75

vortex by Søiland and Rossby (2013) also supported energy transfer via eddy merging

76

mechanism, but also recommended the need to study this in detail.

77

There are two main disagreements among earlier studies. The first is regarding the drift of the

78

vortex. Ivanov and Korablev (1995b) suggested that the vortex drift is guided by the mean

79

cyclonic circulation that largely follows f/H contours as described by Nøst and Isachsen (2003).

80

This was later contradicted by Köhl (2007), who modelled the drift of the Lofoten Vortex to be

81

anticyclonic due to the circulation created by cyclones surrounding the vortex. The second

82

disagreement is regarding the size of the vortex. While Ivanov and Korablev (1995a, b), reported

83

the radius of the vortex to be 15 km, Koszalka et al. (2011) retrieved mean velocities and eddy

84

kinetic energy in the Lofoten Basin from surface drifters and found the vortex radius to be 75

85

km. Recently, Søiland and Rossby (2013) documented the solid body core of the vortex to be of

86

7–8 km radius. The uniqueness of the vortex in the Nordic Seas has not yet been fully

87

documented.

88

This study performs a comprehensive observational based quantitative analysis of the vortex,

89

hereafter termed ‘the Lofoten Vortex’, using a suite of long term hydrographic and satellite

90

observations. The main objectives of this paper are to document the uniqueness of the Lofoten 4

91

Vortex in the Nordic Seas; to quantify the vortex’ surface characteristics on seasonal, inter-

92

annual, and climatological time-scales; to show its influence on regional hydrography on

93

seasonal and climatological time-scales; and to study eddy merger processes in detail.

94

2. Data and Methods

95

We use altimeter data to detect the eddies in the Lofoten Basin, in particular the Lofoten Vortex,

96

and to estimate the radii, eddy kinetic energy, as well as drift velocities. Eddies are tracked and

97

mergers with the vortex and the resulting vorticity changes are quantified. Hydrography is used

98

to describe the signature of the Lofoten Vortex.

99

2.1. Altimeter data

100

High-resolution sea level anomalies (SLA) during the past 16 years (1995–2010) are used to

101

study the vortex. The SLA fields, corrected for the inverse barometer effect, tides, and

102

tropospheric effects (Le Traon and Ogor, 1998), are based on merged Envisat and ERS-I and II

103

altimetric data (Ducet et al., 2000). The SLA fields obtained from AVISO are provided as

104

weekly means on a 1/3° Mercator projection grid. This corresponds to a resolution (twice the

105

grid spacing) of 22–29 km in the Lofoten Basin study region (small box in Fig.1) and 24–27 km

106

in the tracking areas (68°N to 72°N and 4°W to 20°E; Fig. 2). In a recent study, Volkov and

107

Pujol (2012) showed that the root mean square difference between the altimeter derived sea

108

surface height and tide gauge measurements in the Norwegian Sea is generally 3 cm. The SLA

109

fields are used for the detection and quantification of the Lofoten Vortex (discussed in Section

110

2.1.1). Sea surface geostrophic velocity anomaly components, u' and v', are computed from the

111

SLA gradients using the conventional geostrophic relation:

5

112

-g ∂h u'= f ∂y

(1)

113

g ∂h v'= f ∂x ,

(2)

114

where g is the gravitational acceleration, f is the Coriolis parameter, h is SLA, and x and y are the

115

longitudinal and latitudinal coordinates. From the geostrophic velocity anomalies, eddy kinetic

116

energy, EKE, is computed using the standard relation (Chaigneau et al., 2008):

117

EKE=

118

In addition to SLA fields, we also use gridded weekly absolute dynamic topography (ADT) data

119

(1995–2010) in this study. The ADT fields are used to quantify the relation between the

120

seasonality of the slope current, and the strength of Lofoten Vortex, using an along-isobath

121

approach. Since the slope current is a nearly barotropic shelf-edge current (Skagseth et al., 2004)

122

this barotropic estimate of the transport is reasonable. The volume transport (1 Sv=106 m3 s-1) is

123

estimated from ADT using the following equation:

124

gH Ψ (l,t)= f ∆H ADT,

125

where l is the along-slope coordinate, and H is the mean bottom depth of the area between the

126

500 m and 800 m isobaths. Here, ∆H ADT = ADT(H500m, l, t) - ADT(H800m, l, t) and represents the

127

difference in the cross-slope ADT between these two isobaths. The ADT dataset is the sum of

128

the time invariant CNES-CLS09 mean dynamic topography (MDT; Rio et al., 2011) and the time

129

variant weekly SLA data described above. The errors associated with the estimation of CNES-

130

CLS09 MDT are provided together with the MDT dataset (Rio et al., 2011), and in the Lofoten

131

Basin, from the continental plateau and out, the errors are less than 1 cm (not shown). The errors

u'2+v'2 2 .

(3)

(4)

6

132

associated with ADT are the quadratic sum of the errors in SLA (3 cm; Volkov and Pujol, 2012)

133

and MDT (1 cm, Rio et al., 2011), i.e., 3 cm.

134

2.1.1. Automatic eddy detection

135

The detection of Lofoten Basin eddies was done using the automated hybrid algorithm

136

described by Halo (2012). It combines the two most used criteria to identify an eddy, namely,

137

closed contours of streamlines of sea surface height (Chelton et al., 2011) and a negative Okubo-

138

Weiss parameter (W; Isern-Fontanet et al., 2006; Chelton et al., 2007), i.e., where the vorticity of

139

the flow field dominates its deformation (W