Motyka et al LeConte Suppl Mat GRL ver 6

advertisement
1
Supplementary Material
2
Rapid Submarine Melting Driven by Subglacial Discharge, LeConte Glacier, Alaska
3
Authors: Roman J Motyka1,2*, William P Dryer1,2, Jason Amundson2, Martin Truffer1, and Mark
4
Fahnestock1
5
1
6
99775
7
22
8
Juneau, AK 99801
9
*rjmotyka@uas.alaska.edu
Geophysical Institute, University of Alaska Fairbanks; 903 Koyukuk Drive, Fairbanks, AK
Department of Natural Sciences , University of Alaska Southeast; 11120 Glacier Hwy(SOB1),
10
11
1. Heat and Salt Balances
12
13
[1] Determination of water and heat fluxes followed methodology presented in Motyka et al.
14
[2003] and later used by Rignot et al. [2010]. This method utilizes the fact that melting of
15
glacier ice absorbs considerable thermal energy, the source of which is inferred to be incoming
16
warm “ambient” seawater at depth (Figure 2). The benefit of this method is that the details of
17
convection and heat transfer along the face do not have to be known in order to determine the
18
amount of submarine melting. Our analysis focuses on the outflow plume and uses 600 kHz
19
ADCP measurements to depths of 70 m with 2-m bins to evaluate fluxes in the outflow plume.
20
The 2-m-bin data were linearly interpolated vertically to one meter intervals and horizontally
21
between stations to 10 m intervals. Currents for the top 4 m cannot be resolved by the ADCP; we
22
therefore used the data trends from the bins below 4 m to extrapolate to the surface. To
23
determine the depth of the outflow plume (Table S1), we used current, salinity, and temperature
24
profiles. We also used a Gade melt line analyses described in Mortensen et al. [2011] to
25
determine the fraction of submarine glacier meltwater in the water column.
26
27
[2] Spacing between stations was a compromise between temporal and spatial resolution. In
28
order to resolve tidal effects, we strove to complete each transect in 2 hr or less. To provide
29
adequate spatial resolution, we attempted to space our stations about 200 m apart or less. In
30
practice, time per station and distance between stations averaged 2.2 hr and 190 m respectively
31
(Table S2). The number of stations occupied per transect varied from 5 to 7.
32
33
1.1 Salt Balance
34
[3] The total water flux in the outflow plume, Qp, was calculated by numerically integrating the
35
measured, interpolated, and extrapolated currents across the flux gate:
36
Qp = WD ui(x,z) dxdz
37
(1)
38
39
where ui(x,z) are the 600 kHz ADCP velocities, D is the plume depth, and W the width of the
40
flux gate. Some flow along the walls is missed by our stations.
41
42
[4] The water discharge in the outflowing plume, Qp, has both seawater, Qsw, and freshwater,
43
Qfw, components:
44
45
46
Qp = Qsw + Qfw = Qsw + Qsg + Qm
(2)
47
where Qsg and Qm are the subglacial discharge and terminus ice melt, respectively. The fraction
48
of seawater in the overflow plume was determined from the salinity. Subglacial discharge, Qsg,
49
and glacier melt Qm, are assumed to have zero salinity and to be at the pressure melting point.
50
We used Ssw = 28 PSU and Tsw = 6.8 °C for ambient seawater salinity and temperature; these
51
were the average salinity and temperature in the lowest part of the water column (~ 170 – 180
52
m). We estimated the fraction of seawater (s ) in the plume at a depth z from
53
s = Sp(z)/Ss
54
(3)
55
56
where Sp is the measured salinity in the plume. The seawater flux in the outflow plume is then
57
given by
58
Qsw = WD ui(x,z)  si(x,z) dxdz
59
(4)
60
61
Equation (4) was numerically integrated using 600 kHz ADCP velocities, ui(x,z), and Spi(x,z)
62
obtained from CTD casts.
63
64
[5] The freshwater flux in the outflow plume, Qfw, was determined from eqn. 2.
65
66
1.2 Heat Balance
67
[6] From conservation of mass, the incoming deep ambient seawater flux must be equal to the
68
flux of seawater in the outgoing plume and we use Qsw to evaluate Qm. From conservation of
69
energy, the heat coming into the system via the warm seawater, Hsw, and via the subglacial water,
70
Hsg, must equal the heat leaving the system in the plume, Hp, plus the latent heat lost to melting
71
ice, Hm:
72
Hm = H = Hsw + Hsg - Hp
73
(5)
74
75
Here we neglect heat generated by friction and heat lost to the atmosphere, and we assume that
76
Hsg = 0 because the ice and subglacial water are at 0 C, which is our reference state. We
77
determine the incoming heat flow from
78
Hsw = sw Qsw Csw Tsw
79
(6)
80
81
where Csw is the specific heat (3990 J kg-1 K-1), sw is the density (1022 kg m-3), Tsw is the
82
temperature of the deep seawater (6.8 C) and where, from conservation of seawater in the fjord,
83
the rate of seawater inflow near the terminus is assumed to be equal to outflow of seawater in the
84
plume, Qsw, which was calculated above.
85
86
[7] Heat carried away from the terminus by the overflow plume was computed by numerically
87
integrating across the flux gate:
88
89
90
H p = WD ( pi ui C p T pi )dx dz
(7)
91
using the measured and interpolated values of mixed-water density, temperature, and velocity
92
within the plume. The heat capacity of the plume water was taken to be that of seawater, Csw.
93
The rate of ice melt, Qm, is then computed from:
94
Qm = Hm L-1
95
(8)
96
97
where L is the latent heat of fusion for ice. An adjustment for saltwater melting-point depression
98
of approximately -1.0 C was included in this calculation. Qsg is then evaluated from Qfw and Qm
99
(eqn. 2).
100
101
2. Error Analysis
102
103
[8] The SeaCAT temperature and salinity values are highly accurate, ±0.005 °C and ±0.007 PSU
104
respectively; thus the primary sources of uncertainty in our methods stem from current
105
measurements, interpolations, and extrapolations. The accuracy for ADCP current
106
measurements is reported to be 1 cm/s when averaging 50 pings [RDI, pers. comm., 2012]. We
107
preformed a sensitivity analysis on our 600 kHz ADCP data sets, each of which usually consisted
108
of 200+ pings. We divided data for each 2-m bin into two populations, odd and even numbered
109
(or every other ping), and then computed the average velocity, Uodd and Uev, of each data set of
110
100+ pings for each bin. Uodd and Uev for each bin were then differenced for all ADCP casts
111
(1250 total bins). The results produced a normal distribution with a mean of ~ 0 and a standard
112
deviation, σ, of 1.5 cm/s. We interpret this σ to be a measure of variance in currents due to
113
inherent fluctuations. We opted to use 2σ as a conservative estimate of uncertainties for each
114
individual current measurement. Given the normal distribution of the comparison, we consider
115
the error to be random. Therefore, to estimate total uncertainty in flux through the gate, we used
116
σ1 = 2σ√n2, where n is the number of current measurements made across the gate. These
117
uncertainties ranged from 3% to 6.5% of the total flux values. Additional uncertainties accrue
118
from horizontal interpolations and extrapolations. Temperature and salinity varied little
119
horizontally across the transects, so linear interpolation of T and S is unlikely to introduce
120
significant errors. There was variation in currents across the gate; however distances between
121
stations were small enough that we consider linear interpolation to be a reasonable
122
approximation. To be conservative, we added an additional 5% uncertainty to our flux estimates
123
to account for potential uncertainties in our interpolations and extrapolations. These uncertainties
124
apply to the temporal (~2 hr) and spatial (~ 90 m wide swath) averages of each survey.
125
126
Table S1. Characteristics of outflow plume.
127
Transect
Peak
Depth of
Width of
Depth of
Tide
Time local
velocity,
peak vel,
plume,
plume,
stage
cm/s
m
m
m
T3, 9/7
40
27
450
45
↑
T4, 9/7
31
28
40
↑
T5, 9/8
44
13
400
45
─↓
9 - 10.7
T6, 9/8
50
5
400
45
↓─
10.7-12.3
T7, 9/8
45
5
450
45
─↑
12.3 - 14.6
T8, 9/8
52
5
450
45
↑
14.6 - 16.5
T9, 9/9
61
9
450
50
─↓
9 - 11.5
T10, 9/9
58
7
500
50
↓─
11.5 - 14.2
14 - 16.9
T11, 9/10
68
11
500
50
─↓
12.8 - 15.7
128
129
Table S2. Characteristics of transects.
130
Transect
T3
T5
T6
T7
T8
T9
T10
T11
Average
Average
spacing, m
163
228
231
183
196
213
167
179
188
Average Drift,
m
71
93
92
93
71
112
138
112
87
Average time,
hr
2.8
1.8
1.6
2.3
2.0
2.5
2.6
2.8
2.2
Number of
Stations
7
5
5
6
6
6
7
7
131
132
Supplementary Figures
133
Figure S1. Climate data from Petersburg weather station during period of hydrographic surveys
134
at LeConte Glacier and Bay.
135
136
Figure S2. Hydrographic cross-sections for Day 1 surveys. Top panel shows tide with red boxes
137
marking the surveys. Left panels show results of ADCP measurements; the 600 ADCP was not
138
used for first two surveys. Right panels show salinity and temperature cross-sections.
139
Figure S2, con’t. Hydrographic cross-sections for Day 1 surveys, continued. Left panels show
140
turbidity in NTU, log scale. Right panels show percentage melt water and subglacial discharge
141
using methods in Mortensen et al. [2013]. .
142
143
Figure S3. Hydrographic cross-sections for Day 2 surveys. Top panel shows tide with red boxes
144
marking the surveys. Left panels show results of ADCP measurement. Right panels show
145
salinity and temperature cross-sections.
146
Figure S3, con’t. Hydrographic cross-sections for Day 2 surveys, continued. Left panels show
147
turbidity in NTU, log scale. Right panels show percentage melt water and subglacial discharge
148
using methods in Mortensen et al. [2013]. .
149
Figure S4. Hydrographic cross-sections for Day 3 surveys. Top panel shows tide with red boxes
150
marking the surveys. Upper left panels show results of ADCP measurements. Upper right panels
151
show salinity and temperature cross-sections. Lower left panels show turbidity in log scale.
152
Lower right panels show percentage melt water and subglacial discharge using methods in
153
Mortensen et al. [2013].
154
Figure S5. Hydrographic cross-sections for Day 4 survey. Top panel shows tide with red boxes
155
marking the survey. Upper left panels show results of ADCP measurements. Upper right panel
156
shows salinity and temperature cross-sections. Lower left panel show turbidity in log scale.
157
Lower right panel shows percentage melt water and subglacial discharge using methods in
158
Mortensen et al. [2013].
159
160
Supplementary References
161
162
163
Motyka, R., Hunter, L., Echelmeyer, K. & Connor, C., 2003, Submarine melting at the terminus
of a temperate tidewater glacier, Leconte Glacier, Alaska, USA. Ann. Glaciol. 36, 57-65.
164
165
Rignot, E., Koppes M., and Velicogna, I., 2010, Rapid submarine melting of the calving faces of
West Greenland glaciers. Nature Geosci., 3, 187-191.
166
167
Mortensen, J., J. Bendtsen, R. J. Motyka, K. Lennert, M. Truffer, M. Fahnestock, and S.
168
Rysgaard, 2013, On the seasonal freshwater stratification in the proximity of fast-flowing
169
tidewater outlet glaciers in a sub-Arctic sill fjord. J. Geophys. Res. Oceans, 118, 1382–1395,
170
doi:10.1002/jgrc.20134.
Download