GIS5935 M2.2: Interpolation

 Deliverable 8: Corrected tension spline surface

    In this week's lab, we were creating a surface of water quality in Tampa Bay using different interpolation methods to show how different techniques interpret spatial data, and how those choices shape the final map. We explored several interpolation methods to visualize BOD (mg/L) concentrations, each with its own strengths and weaknesses.

    Thiessen polygons offered a simple "zone" feature approach, assigning each area the value of its nearest sample point. While it was easy to construct, this method creates abrupt boundaries and ignores gradual changes in water quality. The IDW (Inverse Distance Weighted) interpolation improved on this by weighing nearby points more heavily, producing a smoother surface that still closely matches the original data, but can underrepresent spatial variation when points are clustered or unevenly spaced.

    Spline interpolation, especially when working with the regularized version, created the most visually continuous surface, but also introduced distortions in areas with sparse data. In one case, a cluster of low-value points nearly overlapped a high-value reading, causing the spline to bend sharply and misrepresent the transition. To correct this, I modified the original dataset by removing one overlapping low-value point, which allowed the tension spline to produce a more realistic gradient.

    Each method reflected a different style of spatial modeling. Thiessen is rigid and categorical, IDW is weighted but seemingly conservative, and spline is smooth but very sensitive it seems. Choosing the right method depends on the data’s distribution and the goal of the analysis. For this lab, the spline interpolation, after careful data adjustment, provided the most spatially representative surface in my opinion.