Improvements in Measuring Unfrozen Water in Frozen Soils Using the Pulsed Nuclear Magnetic Resonance Method

Abstract

The pulsed nuclear magnetic resonance (P-NMR) method is commonly used to determine unfrozen water content in frozen soils. Traditionally, unfrozen water content is calculated using first return data from the free induction decay (FID) signal intensity (SI); however, these data may contain both ice and liquid water responses, thus overestimating unfrozen water. To address this, the authors used the normalization method for calculating unfrozen water. First, the authors refined a P-NMR testing system for greater temperature control. Next, testing using two P-NMR instruments at different facilities yielded repeatable unfrozen water measurements to within 2.6% of each other. Results from a desorption test indicated excellent agreement (with an average difference of 0.33%) between P-NMR-derived unfrozen water and physical gravimetric water content data. Finally, comparing results from the normalization method to the first return data indicated that calculated unfrozen water in a frozen soil is up to 3.3 times greater for a given temperature using the first return data versus the normalization method. Because this difference is significant, the authors recommend further investigation using an independent method to confirm the results.