 |
| Titel |
A rainfall amount weighted meteoric water line for use in hydrological applications |
| VerfasserIn |
C. E. Hughes, J. Crawford |
| Konferenz |
EGU General Assembly 2009
|
| Medientyp |
Artikel
|
| Sprache |
Englisch
|
| Digitales Dokument |
PDF |
| Erschienen |
In: GRA - Volume 11 (2009) |
| Datensatznummer |
250024971
|
|
|
|
|
|
| Zusammenfassung |
| Use of stable isotope data in precipitation from GNIP or other sources commonly involves
determining a local meteoric water line (LMWL) based on a least squares regression
(LSR) of monthly rainfall data. Local meteoric water lines are used in a variety
or hydrological applications, commonly to determine the relationship of surface
or groundwaters to a potential precipitation source or to determine the degree of
evaporative enrichment of the water. The intersection between the LMWL and a
local evaporation line is commonly used as a start point for calculating evaporative
enrichment.
The equations widely used to determine the LMWL give equal weighting to all data
points regardless of the rainfall amount they represent. In reality smaller rainfall amounts are
more likely to have a lower d-excess due to re-evaporation of raindrops or biases in the
sampling method. Larger rainfall events tend to be more depleted in the heavier isotopes. By
allowing small rainfall amounts, which may have experienced evaporative enrichment, to
have equal influence on the slope of the LMWL, a bias towards the less hydrologically
significant rainfall is introduced. For applications of the LMWL relating to groundwater
recharge, dam storage and major flow events it is the higher rainfall events that are
most important, so it is appropriate to use a LMWL that is weighted towards those
events so as not to overestimate the d-excess of hydrologically important depleted
rainfall.
We propose the use of a rainfall amount weighted LMWL (δ2H = a -
δ18O + b) for
hydrological applications, where the parameters a and b in the line of best fit, yi = axi + b,
are obtained by minimising the least squares equation:
-n p (y - ax - b)
i=1-i--i----i---
LS = -n
i=1 pi
where pi is the rainfall (or precipitation) amount, n is the number of measurements and y and
x represent δ2H and δ18O, respectively.
Using the GNIP we test the hypothesis that the rainfall weighted LSR will have larger
slopes for most sites, with the impact most pronounced in arid and semi-arid areas. For Alice
Springs in arid central Australia the ordinary LSR gave a LMWL of δ2H = 6.86-
δ18O + 4.48,
whereas the rainfall weighted LSR was δ2H = 7.52-
δ18O + 9.3. In contrast relatively
small increases in the slope of the LMWL were observed at Australian coastal sites. |
|
|
|
|
|
|