Repeated measurements on permanent plots using local variability sampling for monitoring soil cover

• Published in: Catena (Giessen) Vol. 73; no. 1; pp. 75 - 88
• Main Authors: Wang, G., Gertner, G., Anderson, A.B., Howard, H.
• Format: Journal Article
• Language: English
• Published: Cremlingen-Destedt Elsevier B.V 15.03.2008; Catena
• Subjects: America; Bgi / Prodig; Remote sensing; Sampling design; Soil erosion cover factor; Spatial variability; United States of America; West North Central; United States of America; Kansas; Spatial variability; Sampling design; Remote sensing; Soil erosion cover factor; Soil erosion; Sampling; Spatial variation
• ISSN: 0341-8162; 1872-6887
• DOI: 10.1016/j.catena.2007.09.005

On US military installations, training activities such as vehicle use disturb ground and vegetation cover of landscapes, and increase potential rainfall runoff and soil erosion. In order to sustain training lands, soil erosion is of major concern. Thus there is a need for sampling designs to monitor degradation and recovery of land conditions. Traditionally, permanent plots are used to obtain the change of land conditions. However, the permanent plots often provide less information over time in characterizing the land conditions because of the fixed number and locations of plots. In this paper, we analyzed the sufficiency of a permanent plot sample and developed a method to improve the re-measurements of the permanent plots over time for a monitoring system of soil erosion based on spatial and temporal variability of a random function. We first applied a local variability based sampling method to generate reference samples that have sampling distances varying spatially and temporally to monitor a soil erosion relevant cover factor for an installation, Fort Riley, USA. Then, we compared a permanent sample with the reference samples annually over 13 years to determine additional sampling in the areas with high variability and temporarily suspending measurements of the permanent plots in the areas with low variability. The local variability based sampling provides estimates of local variability of the cover factor and thus is more cost-efficient than random sampling. By comparison with a reference samples, the re-measurements obtained should more accurately characterize the dynamics of the land conditions.