Stable isotope geochemistry of precipitation and surface waters

Laguna Mata Redonda, Costa Rica.

In order to constrain the relationship between δ¹⁸O in speleothem calcite and climate, it is necessary to understand the modern climate-δ¹⁸O relationship in tropical rainfall. To do this, Bill Patterson (U. Saskatchewan) and I are studying both the temporal and spatial variations in stable isotope values (δ¹⁸O and δD) in tropical precipitation and surface waters. Rainfall δ¹⁸O records were produced by the IAEA Global Network of Isotopes in Precipitation, with several stations in Costa Rica (operated for a few years in the 1990s) and one in Panama (operated from 1968 to 1997). Since the IEAE station density is low, and it is very expensive and time consuming to collect monthly rainfall over several years out of country, we are also using surface waters (rivers and lakes) as a proxy for rainfall. The results of are in two papers in/submitted to the Journal of Hydrology: Lachniet and Patterson, 2002 covers the stable isotope values of Costa Rican surface waters, and Lachniet and Patterson (2006) deals with the statistical analysis, using correlation and multiple stepwise regression, of ~160 surface waters from Panama. See publications for details.

We are currently analyzing surface water δ¹⁸O and δD values for the country of Guatemala (2008).

Surface water sample locations for Panama and Costa Rica. The samples were collected over ~7,000 km of driving distance, and an estimated one million pot holes, one flat tire, several liters of guaro, several gallons of antifreeze, and at least two wayward chickens unlucky enough to stray in front of the car while Bill Patterson was behind the wheel.

We collected ~230 river, lake, and rain water samples from Costa Rica (Lachniet and Patterson, 2002) and Panama (Lachniet and Patterson, 2006) which indicate that surface waters are formed in isotopic equilibrium and lie along the meteoric water line of dD = 7.6 δ¹⁸O + 10.1, (r² = 0.97), the same as observed for numerous tropical stations (Gonfiantini et al., 2001). These data span 5° of longitude and 2° of latitude from both Caribbean and Pacific slopes, so the result appears robust for our area.

The δ¹⁸O values of tropical precipitation are determined primarily by the 'amount' effect (Dansgaard, 1964; Rozanski et al., 1993; Araguás-Araguás et al., 1998; Lachniet and Patterson, 2002) and are poorly correlated with ground temperature (Dansgaard, 1964; Fricke and O'Neil, 1999). The 'amount' effect arises from strong vertical convection in the tropical atmosphere. As moisture parcels rise and cool, the precipitation becomes increasingly depleted as the heavier H2O molecules are preferentially removed. The gradual removal of moisture along a storm track and consequent decrease in δ¹⁸O is known as the 'continental' effect and generally operates on large spatial scales such as over the Amazon Basin (Vuille et al., in press) and North America and Europe (Rozanski et al., 1993). Further, the 'orographic' effect occurs when an air mass is lifted, cooled, and rainout produces increasingly depleted values. A change in the δ¹⁸O value of the evaporating oceans will also affect the δ¹⁸O value precipitation, as will a change in moisture source in continental settings (i.e. Pacific Ocean vs. Gulf of Mexico moisture for North America).

Plot showing the 'amount' effect in precipitation in Panama and Costa Rica (monthly means)

δ¹⁸O values of monthly precipitation show a clear amount effect (Lachniet and Patterson, 2002; 2006). δ¹⁸O values (in ‰ VSMOW) are lower during the wet season and higher during the dry season, yielding a slope of -2.62‰/100 mm precipitation (r² = 0.82). Conversely, the correlation between temperature and δ¹⁸O is poor (r² = 0.06). Since El Niño events are associated with decreased precipitation on the Pacific slope of Panama, the δ¹⁸O value of rainfall should be higher during these periods. The short time series of δ¹⁸O in Panama precipitation contains δ¹⁸O data for 6 El Niño events and 5 La Niña events. While the time series is short, it does show that mean δ¹⁸O values are higher during El Niños (-4.9‰) than La Niñas (-5.3%), a finding consistent with ENSO forcing of δ¹⁸O in precipitation.

Plot of the Central America surface water line, containing data from ~230 surface water samples from Costa Rica and Panama.

Plot of d18O vs topography across the Talamanca Mountains, Costa Rica

The δ¹⁸O values show variations that mimic the topography, such as that evident in the plot above, which is a transect from the Caribbean Sea in the east to the Pacific Ocean in the west. The δ¹⁸O values (note inverted scale) show a depletion as they traverse the high Talamanca Mountains, while sea-level rivers are about 2‰ lighter on the Pacific vs the Caribbean Slope. The variations reflect the temperature effect, and perhaps also the rainout effect as air masses traverse the isthmus. Note though, that at any one location, the amount effect is the dominant source of δ¹⁸O variability.

This plot shows the decrease in δ¹⁸O values as moisture masses traverse the Isthmus of Panama from the Caribbean to the Pacific. The decrease in δ¹⁸O values represent rainout associated with orographic uplift of Caribbean air masses. Stepwise multiple regression of δ¹⁸O against various environmental parameters results in an equation that explains 74% of the isotopic variance of surface waters in Panama. The dominant environmental variables are distance from the Caribbean, median stream elevation, and latitude and longitude.