IAEA CRP on 18 O MIBA and Planning

IAEA CRP on 18 O (MIBA) and Planning for Tumbarumba IPILPS Workshop ANSTO 18 -22 April 2005

Overview of talk • Data acquisition - How did we get to where we are today? The MIBA network IAEA and other planning meetings ANSTO’s contribution Why Tumbarumba? Sampling a priori estimates and our working hypothesis

The MIBA network • Moisture Isotopes in the Biosphere and Atmosphere • Initiated at a meeting in Vienna, May 2004 • Building on GNIP (& GNIR) in collaboration with WMO • Primary aim - to facilitate acquisition of data on stable isotopes in biospheric and atmospheric water.

Moisture Isotopes in the Biosphere and Atmosphere (MIBA) IAEA Idealized distribution- 100 sites: 90 continental, 10 oceanic (water vapor) 15 15 15

Moisture Isotopes in the Biosphere and Atmosphere (MIBA)

The MIBA network • Objectives 1 Regional scale hydrological budgets; 2 The partitioning of annual carbon fluxes; 3 The development of new global change indicators; 4 Ecosystem functioning; 5 Interpretations of 13 C and 18 O analyses in organic matter; 6 The validation of general circulation models; and 7 Past global responses to climate change.

The IAEA CRP • Title: Isotope methods for the study of water and carbon cycle dynamics in the atmosphere and biosphere • 1 st RCM this May in Vienna

The MIBA network • IAEA Membership P. Aggarwal, Head, Isotope Hydrology Section IAEA D, Yakir (Israel; Chair), G. Farquhar (Australia), L. Flanagan (Canada), F. Longstaffe (Canada), R. Siegwolf (Switzerland), G. Hoffman (France), H. Meijer (the Netherlands), H. Griffiths (UK), J. Berry (USA). P. Tans (USA), P. Ciais (France), N. Buchmann (Switzerland), L. Sternberg (USA), T. Dawson (USA), G. Lin (China), W. Stichler (Germany), J. White (USA), and J. Santruek (Czech Republic), Brent Helliker (USA), A. Henderson-Sellers (Australia).

The IAEA CRP - the Australian contribution • Routinely measure SI in Canberra vapour. • Compare fortnightly 18 O in grasses, trees, soil-water, rain and vapour near Canberra. • Monitor 18 O less frequently in grasses, trees, soil-water, rain and vapour at sites in the MDB & use these data to evaluate climate models and for model intercomparisons. • Measure SI and chemistry in monthly rain-water samples across the continent. • Characterisation of 18 O in ground-water in relation to soil properties. • Refine MDB water balance models using the data above.

ANSTO’s contribution to the CRP • IPILPS • Water isotopologues 1 H 16 O; 1 H 18 O 2 2 and 1 H 2 H 16 O • Comparison of LSS By output, and against real data • Real field data required at appropriate time scales • Not much exists….

Initial site selection • Ozflux sites in Murray Darling Basin (MDB) Padthaway

Tumbarumba - Oz-flux tower • CSIRO Division of Atmospheric Research, Dr. R. Leuning & Dr. H. Cleugh

Tumbarumba • • Bago Forest. 35ºS, 148ºE 1200 m Cool-temperate zone, MDB Dominant vegetation Eucalyptus delegatensis (Alpine Ash) with E. dalrympleana (Mountain Gum)

Tumbarumba

Tumbarumba • Routine measurements comprise: temperature, humidity (bulk concentration of water vapour), wind (speed and direction), net radiation (both shortwave and longwave components), surface pressure, soil moisture and temp as well as the fluxes of (bulk) water vapour and heat.

Tumbarumba - other major collaborators • University of Wollongong, Prof D. Griffith portable FTIR MS • ANU, Prof G. Farquhar & Dr C. Keitel; Dr H. Stuart-Williams leader of the IAEA CRP and of a research team interested in C and O isotope fractionation applications in plant physiology. Stable isotope analysis. • CSIRO Forestry, Dr H. Keith primary interested in forest productivity • CSIRO Land & Water, Dr A. Herczeg & Dr F. Leany; Dr J. Deighton GNIP. Stable isotope analysis

What has ANSTO sampled? • Precipitation, dew, surface soil, deeper soil, tree stem, leaves and vapour at various heights above the ground, sampling most types hourly during daylight and less frequently at night over 5 days in early March 05. • We will infer from our results § surface water vapour fractionation (from soil-surface water values, using the Craig & Gordon model) § equilibrium transpired vapour and root zone water fractionation (from stem water) § non-equilibrium transpired vapour (from total flux and leaf water)

Sample preparation and analysis • All samples were sealed and frozen or chilled in the field ASAP after collection. • Vegetation samples will be vacuum distilled at ANSTO to collect unfractionated water (~ 2 hrs per sample). • Plant distillates, precipitation, dew and vapour condensates are being analysed for SWIs at ANU. • A pyrolysis method is used. 2 H <1 o/oo and 18 O <0. 3 o/ oo precision.

Sample preparation and analysis • Soil waters are being analysed by CSIRO Land & Water in Adelaide following azeotropic distillation using kerosene to prevent fractionation. With CO 2 equilibration the precision is <0. 05 o/oo for 18 O • Total soil moisture has been measured at ANSTO by drying to constant weight. This will be followed by particle size analysis for soil characterisation.

Sampling - Precipitation • Event-based precipitation was collected at ground level, using collection devices that ANSTO built according to an IAEA (2003) design • Dew was collected several mornings from a plastic sheet laid on the ground

a priori sample estimates - Precipitation • Should lie on the GMWL d 2 H = 8 x d 18 O + 10 o/oo but will vary due to local conditions, particularly distance from the source & altitude Estimates for Tumbarumba at 1300 m altitude using an on-line calculator range from: -4. 8 to -7. 2 o/oo (d 18 O) and -25 to -50 o/oo (d 2 H) with averages of -6. 5 and -39 o/oo • http: //www. waterisotopes. org/

Sampling - Vapour • Lagged sample tubes were installed on the tower to nine heights (separate matched sets for our vapour & Uo. W’s FTIR) • Filtered to prevent contaminants and blockage

Sampling - Vapour • Vapour tubes were attached to ANSTO cold traps amended from Helliker et al (2002) design. • Samples were collected hourly during daylight. • Lower frequency at night due to lower vapour pressures

Sampling - Vapour • Last minute amendment based on advice from ANU. • Installed a ground level sampler with a lid collecting input air at 10 m • Lid removed between FTIR cycles to avoid humidity and temperature effects and replaced to allow signal to re-establish

a priori sample estimates - vapour • Atmospheric vapour will be depleted with respect to precipitation depending upon the temperature (and hence altitude). Majoube et al 1971 • Soil vapour will be depleted along LEL compared to the soil water but will become more enriched and approach LMWL as time passes after rain (? Equilibrium? ).

a priori sample estimates - soil vapour • How will soil vapour change over time following rain? Percent H 2 O in surface soil will decline due to surface evaporation until it reaches a steady state due to replacement from deeper water by capillary action Vapour from surface soil will initially be primarily affected by phase change (equilibrium fractionation). It will then enrich due to kinetic fractionation as the proportion of heavy isotopes remaining in the surface water becomes larger. In addition, as the surface gets drier, diffusive fractionation will also come into play as vapour is formed and migrates from lower down the soil profile. • Help from the audience on understanding this process more clearly would be appreciated

Sampling - soil water • Surface scoops ( 5 cm incl litter) were collected at the same time as, and adjacent to, leaf and stems samples. At hourly intervals during daylight. • Cores taken 4 times per day to a depth of 1 m. Samples were split for depth profile and then measured for moisture content or sent for SWI analysis.

a priori sample estimates • Soil water Surface water will reflect any recent rainfall. It will become more enriched over time due to evaporation. It should lie on the LEL. Deeper samples will reflect heavy or longer rainfall and will be biased towards the depleted end of the LMWL range at the intercept with the LEL and will not be enriched. It is assumed this will be well mixed, not change dramatically with incident rainfall (eg Neal & Rosier 1990), and be equivalent to local seeps and xylem water (initial estimates -6 o/oo d 18 O and -40 o/oo d 2 H).

Sampling - xylem Bark (phloem) must be removed to avoid contamination with enriched leaf water and photosynthates. The exposed end is cut off by >2 cm to minimise the potential for exchange with atmospheric vapour. The sample is cut into 1 cm sections to enhance the ease of vacuum distillation. The stem is cut directly into a 12 m. L ‘Exetainer’

$a priori sample estimates - xylem • Should be unfractionated with respect to the$

a priori sample estimates - xylem • Should be unfractionated with respect to the root zone (deeper) soil water (estimates -6 o/oo d 18 O and -40 o/oo d 2 H). • The height of the sample should not matter. Hence, samples from the under story are OK. • At equilibrium, should equal transpired water vapour (but this may only occur for a few hours each day).

Sampling - Leaf water • Leaf water Leaves in sunlight were preferred. An estimate of the hours of sunlight exposure is made. Samples were quickly placed into ‘Exetainers’ Main vein was removed. Air temperature and cloud cover estimates were also made.

a priori sample estimates - leaf water • Will vary considerably diurnally due to stomatal opening (light, time and water potential dependent) and local climatic conditions including temperature, humidity, wind speed and incident radiation. Pre-dawn, should be equivalent to xylem (~ -6 o/oo d 18 O and ~ -40 o/oo d 2 H. Morning, should start to enrich to a maximum around noon in the order of 15 -20 o/oo (D 18 O), 60 -80 (D 2 H). Afternoon, should plateau but then start to return to xylem values as stomata close. Evening, relax back to xylem signature.

a priori sample estimates • Transpired water At equilibrium, it is assumed that transpirate will be equivalent to xylem/source water. Due to the lag induced by the enrichment of the leaves in the morning and the relaxation of that enrichment in the afternoon, transpirate will be more depleted in the morning and more enriched in the afternoon (eg Harwood et al 1998). The degree of difference will depend on the overall water flux rates.

Diurnal variation of (a) the 18 O of transpired water vapour for leaves on day 1 ( ) and day 2 ( , , ) indicating the vapour pressure deficit (VPD) status and general trend over the day (solid line). (b) Evaporative site enrichment (de) for different leaves. Solid lines calculated using H 2 18 O of transpired vapour. Dashed lines represent the trendline fitted to the same data points assuming isotopic steady state (ISS) held throughout the day. Harwood et al 1998 Plant, Cell & Environment 21 (3) 269 -283)

a priori inferred estimates • Vapour from soil water Should be depleted with respect to surface soil water according to Craig & Gordon (1965) model (in the order of -10 to -12 o/oo (D 18 O), -40 to -50 o/oo (D 2 H)). Should lie on the LEL. After rainfall it should become less depleted as the surface soil water becomes more enriched and it should approach the LMWL.

What is the data to be used for? • A Evaluation of LSS output • B Describing a three-point mixing model of dualisotope fractionation in vapour from the atmosphere, soil evaporation and plant transpiration at hourly intervals in daylight hours and less frequently overnight.

L W rainwater M G Surface water Root zone water = stem water = transpired water Surface evaporate L LE Local vapour