This week, I worked on creating surfaces of water quality in Tampa Bay using different interpolation methods. Interpolation helps estimate values at unsampled locations based on known sample points, which is key for understanding spatial patterns in environmental data like Biochemical Oxygen Demand (BOD) concentrations.
I compared four methods: Non-spatial averaging, Thiessen polygons, Inverse Distance Weighting (IDW), and Spline interpolation. Non-spatial simply averages all sample values without considering location, so it lacks any spatial detail. Thiessen divides the area into polygons where each sample controls its own zone, assuming sharp boundaries that don’t usually exist in natural systems. IDW improves on this by weighting nearby points more heavily, producing a smoother surface that respects distance relationships. Finally, Spline fits a smooth surface passing exactly through the points, but it can exaggerate values in areas with sparse data, sometimes creating unrealistic peaks.
Overall, IDW gave the most balanced and realistic surface for Tampa Bay’s water quality, avoiding extremes while capturing spatial trends. Figure 1 shows the IDW surface I generated for BOD concentrations.
No comments:
Post a Comment