Incorporating landscape characteristics in a distance metric for interpolating between observations of stream water chemistry

Abstract

Abstract. Spatial patterns of water chemistry along stream networks can be quantified using synoptic or "snapshot" sampling. The basic idea is to sample stream water at many points over a relatively short period of time. Even for intense sampling campaigns, the number of sample points is limited and interpolation methods, like kriging, are commonly used to produce continuous maps of water chemistry based on the point observations from the synoptic sampling. Interpolated concentrations are influenced heavily by how distance between points along the stream network is defined. In this study, we investigate different ways to define distance and test these based on data from a snapshot sampling campaign in a 37-km 2 watershed in the Catskill Mountains region (New York State). Three distance definitions (or metrics) were compared: Euclidean or straight-line distance, in-stream distance, and in-stream distance adjusted according characteristics of the local contributing area, i.e., an adjusted in-stream distance. Using the adjusted distance metric resulted in a lower cross-validation error of the interpolated concentrations, i.e., a better agreement of kriging results with measurements, than the other distance definitions. The adjusted distance metric can also be used in an exploratory manner to test which landscape characteristics are most influential for the spatial patterns of stream water chemistry and, thus, to target future investigations to gain process-based understanding of in-stream chemistry dynamics.

Using landscape characteristics to define an adjusted distance metric for improving kriging interpolations

“…The combination of landscape characteristic and ω that resulted in the 24 lowest leave-one-out cross-validation K RMSE for each property was selected as the best 1 performing adjusted distance metric for interpolating that particular property. This is similar 2 to the methodology outlined by Lyon et al (2008). 3 4…”

“…While there are conceptual similarities between this current study and the previous 12 work by Lyon et al (2008), it should be noted that the current study considers the local 13 landscape position influence on how samples collected in the landscape relate to each other. 14 This differs from Lyon et al (2008) where the focus is on adjusting path-restricted, in-stream 15 distances between stream water sampling positions based on the average composition of the 16 drainage area to a given sampling point. The current study interpolates point observations of 17 different properties in both the mineral and organic soil layers at the Ovanmyra field site in 18 central Sweden (Figure 1) using both a traditional distance metric and the adjusted distance 19 metric (Eq.…”

“…Based on a traditional Euclidean 11 distance metric, the points are separated by a distance defined simply by a straight-line path 12 (d ij ) based solely on the coordinates of the points. Lyon et al (2008) …”

“…Research is needed on how to define 3 such a weighting value a priori; however, we can take a methodical approach to identify the 4 best possible weighting value for a given landscape characteristic based on a given set of 5 observations. This is similar to the method used in Lyon et al (2008). We tested each of the 6 five landscape characteristic with values for w varying from zero to one in steps of 0.1 (here, 7 we let ω step by increments of 0.1 to limit the number of possibilities).…”

Comparison of hybrid regression and multivariate regression in the regional flood frequency analysis: A case study in Khorasan Razavi province

“…Speyrer and Ragas examined the impact of flood risk and mandatory flood insurance on property values (24). Several studies performed the popular SPARROW model using a hybrid approach combining conventional regression methods with spatial data based on landscape characteristics and stream properties to predict continuous water quality from point observations (25)(26)(27). Firat compared different models for daily river flow forecasting and concluded that adaptive neuro-fuzzy inference system (ANFIS) has higher accuracy and reliability than other models (28).…”