When HEC-MetVue computes basin averages, it is actually computing a volume based on the intersection of the TIN and the polygon in question and then applying this volume to either the entire polygon area or the intersected area of the polygon and defined TIN. Grids cells that are missing do not contribute to either the volume or the intersected area applied to the polygon. A normal TIN that does not represent a grid cannot have missing cells since it is a representation of existing data with any missing data excluded from the triangulation. HEC-MetVue can display averages for any type of data but only displays volumes if the measurement value is in a unit of length, such as inches.
An example of a dataset where it is more appropriate to use the intersected area for a computation would be a gridded dataset that represented snow depth having the grid cells over water bodies set to missing. If the objective is to compute the average snow depth over land, the intersected area between the defined portions of the TIN and the polygon will give the desired value. Below is an example of a TIN with missing cells (blue mesh) intersecting a polygon and displaying both intersected and total area averages and their associated areas. Note that HEC-MetVue will only display both averages and areas if both types are selected in the Properties Window->Layer Settings->Polygon/Basin Averages section and the intersected and total areas for a polygon actually differ. As an alternative to using intersected area, a shapefile with the water bodies represented as 'holes' within the polygon should give nearly the same result as long as the cell resolution can closely approximate the water body area.
When using observer data for a TIN (especially a historical storm), it is typical for the storm to be defined by observer reports. However, the area directly adjacent to the storm would not have observer reports, even though in actuality the area just outside the storm had zero precipitation. Although the TIN does not encompass the zero precipitation area since no reports actually exist, zero precipitation over areas not covered by the TIN is implied. In this case, the correct area average for a polygon is the total polygon area. Below is an example of a TIN where it is more appropriate to use the polygon total area for the computation. In this example, the observer network reports stop short of covering entire areas of subbasins in the eastern side of the watershed as there were no reports there just outside the limits of the storm. Therefore, it would more appropriate in this case to rely on the total polygon area average, which uses a fair assumption of zero precipitation over the portions of the subbasins outside the observer TIN limits.
