Monitoring for Temporal Changes in Soil Salinity using Electromagnetic Induction Techniques

Abstract

Electromagnetic induction surveys are often used in practice to estimate field‐scale soil salinity patterns, and to infer changing salinity conditions with time. We developed a statistical monitoring strategy that uses electromagnetic induction data and repetitive soil sampling to measure changing soil salinity conditions. This monitoring approach requires (i) the estimation of a conditional regression model that is capable of predicting soil salinity from electromagnetic (EM) survey data, and (ii) the acquisition of new soil samples at two or more previously established survey sites, so that formal tests can be made on the differences between the predicted and observed salinity levels. We examined two test statistics in detail: a test for detecting dynamic spatial variation in the new salinity pattern and a test for detecting a change in the field median salinity level with time. We applied this monitoring and testing strategy to two EM survey‐soil salinity data sets collected at multiple points in time from the saline irrigation district of Flumen, Spain, Our results demonstrate that this monitoring approach was successfully able to quantify the temporal changes in the soil salinity patterns occurring within these two fields.

Electromagnetic induction as a basis for soil salinity monitoring within a Mediterranean irrigation district

“…This was the case for the Ebro valley (Amezketa, 2006) and specifically for the Flumen irrigation district (López-Bruna and Herrero, 1996;Lesch et al, 1998;Herrero et al, 2003). The EM38 successfully mapped soil salinity of small 75 plots (< 1 ha).…”

“…Thus, it is very improbable for the salinity to proportionally change temporally, a condition required (Lesch et al, 1998) to maintain the calibration equations between two sampling dates. …”

Statistical Methods for the Analysis of Soil Spatial and Temporal Variability

“…Lesch et al (1998) discussed another approach that can test for change in the mean value as well as for a dynamic spatial variation between two time instants. Dynamic spatial variation refers to the spatially variable change of a soil property between two time instants, i.e., the change is not the same between two time instants for each location.…”

“…To test for the existence of dynamic spatial variation, Lesch et al (1998) identified different steps. First of all, the pairs of soil data obtained at a first time instant (y and z), from n space locations, are used to establish a calibration equation based on the regression model:…”

“…Therefore, in this chapter, we will present the main statistical methods with a short description of each of them, discuss eventually the possible links between some of them, identify their assumptions, merits and limitations, stress some of their practical applications, and list some relevant bibliographic references. The chapter fills the gap from the Heuvelink & Webster (2001) paper by considering more statistical methods like temporal stability (Douaik et al, 2006;Vachaud et al, 1985) and dynamic spatial variation (Douaik et al, 2007;Lesch et al, 1998), updates the presentation of the methods from the Kyriakidis & Journel (1999) paper by taking into account the research work done, and especially the last developments in pedometrics, during the past decade and adds some examples of practical applications like mapping using the Bayesian maximum entropy (Christakos, 2000;Douaik et al, 2005;. After presenting some classical statistical methods (Section 2), we provide an overview of the geostatistical methods including structural analysis (Section 3) and prediction (Section 4).…”