Isabella Velicogna

1
ICE MASS BALANCE AND SNOW ACCUMULATION FROM GRACE
Isabella Velicogna John Wahr Dept. of Physics
C.I.R.E.S., University of Colorado,USA Isabella
_at_colorado.edu
2
OUTLINE Mass Balance, Ice Volume and Snow
Accumulation Estimate From GRACE Satellite
Gravity -Using time variable gravity -Using
along track static geoid -SOME PRELIMINARY
RESULTS
3
When Trying to recover Ice Mass Balance (IMB)
from the GRACE data we can proceed in different
ways 1- use GRACE data Alone gta 1st order
approximation can be to map the change in
mass distribution over a major ice sheet (e.g.
Antarctica or Greenland) and assume that the
dominant contribution to the secular
component of the signal is generated by IMB. --gt
HOWEVER, doing so we are likely to generate an
inaccurate estimate of IMB because GRACE provides
an integral measure of the mass change and it is
not able to separate signals associated with
different processes. 2- Combine GRACE data with
other datasets gt Combination with other
datasets will permit separation of the
signals associated with different processes
4
WHY DO WE WANT TO COMBINE DIFFERENT DATA SETS?
Degree Amplitude function of l contribution
to RMS of geoid from all terms of degree
l 1/2 wavelength20000/l
5
GRACE /Altimeter 2 observations ?
underdetermined Height change ? 3 UNKNOWNS 1 -
IMB 2 - PGR 3 - COMPACTION TREND secular
component of time varying density of the
snow/ice column. GRACE /Altimeter /GPS 3
observations ? can calculate compaction trend
from ?(PGR)0.31 ?(comp) GRACE /Altimeter/ GPS /
ICE CORE MEASUREMENTS 4 observations ? can
calculate compaction trend directly from ice
core ? can reduce pgr error using ?(comp) ?
(PGR)/0.31 NEXT we will add radar-derived
estimates of snow density, and estimates
of height change from radar and In-SAR .
6
Altimeter, GPS Ice cores measurements
CRYOSAT
ICESat/GLAS
7
SIMULATION GRACE -ALTIMETER Simulate 5-yr of
monthly geoid/elevation data 1- Effects of
variable accumulation NCAR CSM-1precipitation,
220 yr trend removed, convolved with
Winghams transfer function. 2- Spatially varying
century-scale ice mass trend (Bently Giovinetto,
1991). 3- PGR signal. 4- Simulated measurements
errors. 5- GRACE geoid contributions from global
redistribution of Mass in the oceans, in
non-polar continental water storage in the
atmosphere. GPS (5-yr simulated daily
coordinates) 1- PGR signal 2- Elastic Earths
response to loading by precipitation and
melting. 3- 1.5 cm RMS random noise on daily GPS
vertical coordinates. Ice Core (5-yr simulated
data) 1- effect of time variable accumulation 2-
10 RMS error in recovering.
8
RMS input 28.4 mm/yr
RMS error 19.9 mm/yr Recover variance of the
input signal within 41
RMS error 17.1 mm/yr Recover variance of the
input signal within 30 gt improvement of 35
9
RMS input 28.4 mm/yr
AND if we add
Precipitation data The if we assume to know the
precipitation field to within 30 error during
the 5 year of the GRACE mission the variance of
the error in the estimate is reduced by and
extra 23.
10
Work in Progress 1--Find optimal ways to solve
for the Compaction error Possible
Solution Combine available datasets to help
constrain Compaction (e.g.gt precipitation data
gt radar/In-Sar derived
estimates of snow
density 2-- Determine IMB for single drainage
basins, How small can we go?
11
GRACE ground tracks. Red 64 days of data.
Blue 1 day of data.
STATIC FIELD (in collaboration with JPL GRACE
Team and LDEO) --Generate high accuracy static
geoid possibly including wavelength down to
100km.maybe smaller? --Development of a
co-kriged optimal interpolation of subglacial
topography over the Antarctic continent using
the high accuracy geoid as a constrain --Ice
mass volume determination from subglacial
topography and ice topography from altimeter data
High frequency content of the K-Band Range dual
one range measurements. Range anomaly after all
signals with more than 30 sec period (200 km
wavelength) have been removed.
12
Monitoring Time variable accumulation
13
Monitoring Time Variable Accumulation From GRACE
14
Monitoring Time Variable Accumulation From GRACE
15
Monitoring Time Variable Accumulation From GRACE
16
Monitoring Time Variable Accumulation From GRACE
17
First Results From GRACE
18
First Results From GRACE