Hi Leela

My first thought was also to wonder about the regression.  What is your range of values and the r-squared value?  (goodness of fit). And are there outliers at either extreme?  They can have a big effect. 

Warm regards
  -Mark Schurr

from my phone

On Sun, Sep 15, 2019, 2:25 PM Tom Millican <[log in to unmask]> wrote:
Hi Leela,
If the values of your samples are confined to a narrow range, compared to the range of what falls in that watershed over the course of a year, it could result in odd slopes with a small sample set. Is the range of values you obtained wide enough?
If snowmelt is involved, it could have an effect on your slope as well.
This is intriguing. I'll be watching for the water isotope experts' ideas.
-tom millican

On Sun, 15 Sep 2019 10:56:34 -0400
Leela <[log in to unmask]> wrote:

> I measured several ground water samples from a temperate region (Lat 42 deg). The hydrogen vs oxygen gives a slope of 3.5 and an intercept of -5. I can understand such a behavior from an arid region. Many samples from an nearby filed gave a slope of 7.7 and intercept of 11. Any suggestions