PRINCIPLES of STABLE ISOTOPE METEOROLOGY

1
PRINCIPLES of STABLE ISOTOPE METEOROLOGY
ABSTRACT The heavy isotopes of water (HDO and
H218O) act as tracers of atmospheric processes
and motions. Three field experiments (Puerto
Escondido, Mexico, KWAJEX, and CAMEX 4) show
that 1 Rain systems with deep, extensive clouds
and organized, closed circulation (e. g.,
hurricanes) produce rain and with heavy isotopes
in concentrations that are markedly lower than in
the surrounding atmosphere. 2 Low-level outflow
from a storm limits the decrease of heavy isotope
concentrations in the vapor and rain, but
transports isotopically depleted vapor to the
surrounding atmosphere.
2
H218O Concentrations and d Terminology
Oxygen-18 is a rare species in water. H218OH2O
.002
-60
Ice Crystals Jet Stream Snow South Pole First
Snow Blizzard Rain Hurricane Eyewall Rain
Hurricane Mean Rain Summer and Tropics Ocean
Water
d terminology gives H218O concentration of sample
(RSAMPLE) relative to that of standard mean ocean
water (RSMOW) in ppt. Most natural waters have
negative ds (lower H218O concentration than sea
water)
d18O (ppt)
0
3
Fractionation
H2O
During condensation heavy isotopes, HDO and H218O
concentrate in precipitation and are
preferentially removed from atmosphere. Result
Heavy isotope concentrations of rain, snow, and
vapor decrease with height
H218O
Rising Air
PRECIP
VAPOR
Rayleigh distillation is the benchmark for
fractionation processes. It assumes that all
water and ice are removed from the atmosphere
immediately upon condensation.
4
Diffusive Equilibration
Early Rain
Later Rain
Falling rain acquires heavy isotopes from the
ambient vapor by diffusion and removes them from
the air. Result Heavy isotope concentrations of
rain and vapor decrease as rain accumulates and
downwind.
Vapor
Vapor
H218O
H218O
Height
H218O
H218O
5
d18O and STORM STRUCTURE
Because heavy isotope concentrations decrease
with altitude, with accumulated rain, and
downwind inside the rain shield, patterns of d18O
reflect major features of storm structure such as
1 storm size and duration, 2 cloud thickness,
3 mean condensation height, 4 low-level
convergence and outflow. Back trajectories show
how much precipitation air has passed through.
They therefore indicate the expected degree of
depression of d18O values below a Rayleigh
distillation model. Typical spatial distributions
of d18O values of rain in hurricanes and
extratropical cyclones are shown below.
6
Extratropical Storm with Overrunning
d18O decreases poleward due to overrunning until
near polar edge where rain acquires 18O from
ambient vapor.
Warm (-5) (-20)
(-10) Cold
Hurricane
d18O decreases inward until near eyewall where
sea spray in high wind restores 18O.
(-2) (-15) (-8) Eye (-8) (-15)
(-2)
7

d18O values of Vapor in Hurricanes are depressed
below Rayleigh distillation curve
8
d18O of Vapor from Three Field Experiments
Key West Clear
Mexico Clear
d18O vs specific humidity during three field
experiments. In placid weather far from
organized storms, d18O values increase as
specific humidity increases. In or downwind from
organized storms, d18O values are much lower
since 18O has been removed
Kwajalein Clear
Key West TS
Kwajalein ITCZ
Mexico TS
Highest d18O values occurred in CAMEX 4 with
almost no organized storm activity. Intermediate
values occurred during KWAJEX with loosely
organized disturbances on ITCZ. Lowest values
occurred as tropical depressions and storms
brushed Puerto Escondido, Mexico
9
HURRICANE FLOYD 14-17 September 1999
1. First rains at any location had high 18O
concentration due to evaporation and acquisition
from unprocessed vapor. 2. 18O concentration of
rain rapidly decreased to first minimum in
association with elevated frontal surface. 3.
18O concentration increased as frontal surface
lowered. 4 . A final period of extremely low 18O
concentration occurred west of storm center due
to two factors A modest lift of frontal surface,
B removal of heavy isotopes by rain upwind.
10
Floyd asymmetric so 18O decreases west of center.
Floyd symmetric so 18O increases west of center.
11
Tracing d18O of VaporWith Back Trajectories
To the west of the hurricane center, the wind and
trajectories (red arrow in the radar image below)
came from the north. Since they passed through a
large area of steady rain before reaching New
Jersey, d18O values were lowered. To the east of
the hurricane center, the wind and trajectories
(white arrow in image below) came from south.
Since they only passed through spotty convection,
d18O remained high. Trajectories shown below and
cross sections above were produced by running the
Meteorological Communitys Mesoscale model MM5 to
simulate Hurricane Floyd.
12
Radar Image 2140 UTC 16 Sept 1999
6
9
13
PUERTO ESCONDIDO, MEXICO 10 - 31 July 1998
Puerto Escondido, Mexico is located in a region
of active tropical cyclogenesis. During period
10-31 July 1998, two tropical storms and one
tropical depression brushed the Pacific coast of
Mexico leaving wakes of vapor with anomalously
low concentrations of heavy isotopes. Between
stormy periods isotope ratios increased to normal
tropical values. Regime changed from stormy
period with generally low d18O values from 10-18
July to quiescent period with high d18O values
from 18 to 31 July. This change of regime was
marked by major change in air trajectories.
14
Tropical Cyclogenesis and ?18O
-10 -15 -25 -30
0 2 4 6 8
Organized Tropical Depressions Produce Vapor and
Rain with low ?18O Leave Vapor Wake with low ?18O
?18O (ppt)
Area of Cloud Tops lt -50oC ()
?18O of Vapor Puerto Escondido July 1998
TMI Images
10 15 20
25 30 Day in
July 1998
Organized System
10 July
Quiescent Period
Incipient System
21 July
28 July
15
Tracing d18O of VaporWith Back Trajectories
Back trajectories and infrared satellite images
below are shown for two different regimes 1. Low
d18O regime 15 - 16 July 1998. Tropical
cyclogenesis with organized precipitation. Back
trajectory passed through the stormy region
before reaching Puerto Escondido. 2. High d18O
regime. 22 - 23 July 1998. Widely scattered
convection. Back trajectory passed through mostly
clear air before reaching Puerto Escondido
16
24 hour back trajectory to 00 UTC 16 July
1998 with organized convection 18 UTC 15 July 1998
24 hour back trajectory to 00 UTC 23 July
1998 with disorganized convection 18 UTC 22 July
1998
17
KWAJALEIN EXPERIMENT 09August - 09 September 1999
KWAJEX was conducted at the northern edge of the
Inter Tropical Convergence Zone (ITCZ). Several
mesoscale wavelike disturbances developed during
this time. Organized, closed circulation did not
develop in any of these systems. Instead, much
vapor was flushed by outflow at low levels.
Consequently, while d18O values were depressed,
they never attained the extremely low values
associated with tropical storms and
hurricanes. The slides below show the poorly
organized structure of the systems and the low
level outflow as indicated by back trajectories.
18
d18O During KWAJEXTracer of Organized Convection

• Vapor with low d18O
• Created in Rain Region
• Advected Downwind
• Persists After Rain Ends

190 T (K) 280






224.75 225.75
226.75 227.75 228.75
229.75
Before Event
Event Upwind
Event Forms
In Event
In Wake
After Event
19
Tracing d18O of VaporWith Back Trajectories

• Trajectory Ending
• 14 Aug 18 UTC
• Remained in Clear Air
• Has High d18O -14

• Trajectory Ending
• 16 Aug 18 UTC
• Passed Under Rain Area
• Has Low d18O -19

1
2
3
3
2
1
224.75 225.75
226.75 227.75 228.75
1
224.75
225.75
227.75
2
3
228.75
226.75
2
1
228.75
3
190 T (K) 280
20

REFERENCES
Gedzelman, S. D., and J. R. Lawrence, 1982 The
isotopic composition of cyclonic precipitation.
J. Appl. Meteor., 21, 1385-1404. Gedzelman, S.
D. and Arnold R. 1994 Modeling the isotopic
composition of precipitation. J. Geophys. Res.
99, 10455-10571. Gedzelman, S. D, .et. al.,
2003 Probing hurricanes with stable isotopes of
rain and water vapor. Mon. Wea. Rev., accepted.
Lawrence J. R., and Gedzelman S. D. 1996. Low
stable isotope ratios of tropical cyclone rains.
Geophys. Res Lett. 23, 527-553. Lawrence, J. R.,
Gedzelman, S. D., Zhang, Z. and Arnold, R. 1998.
Stable isotope ratios of rain and vapor in 1995
hurricanes. J. Geophys. Res. 103, D10,
11381-11400. Lawrence, J. R., S. D. Gedzelman,
J. Gamche, and M. Black 2002 Stable isotope
ratios Hurricane Olivia. J. Atmos. Chem., 41,
67-82. Lawrence, J. R., and S. D. Gedzelman
2003 Tropical ice core isotopes Do they reflect
changes in storm activity? Geophys. Res Lett. 30,
in press.