Some netCDF or HDF datasets store their geolocation as irregularly spaced arrays of pixels or point data. When adding these datasets to a mosaic dataset, the interpolate irregular data function takes the irregularly gridded data and resamples it so each pixel is of uniform size and is square.
When adding variables from netCDF or HDF to a mosaic dataset, it will automatically verify if the data is arrayed regularly. If it is not, the interpolate irregular data function can be used to convert the irregular data into a regularly gridded raster. You can change the interpolation method and cell size used in the interpolate irregular data raster function. For regular-spaced raster data, no interpolation will be applied and the data will be read as it is.
Value Field—If you select a point feature class as the input, you will need to identify the field in the attribute table with the value of the points.
There are four resampling methods for this function:
- Nearest Neighbor—Calculates pixel value using the nearest pixel. If no source pixel exists, then no new pixel can be created in the output.
- Linear tinning—Uses a triangular irregular network from the center points of each cell in the irregular raster to interpolate a surface which is then converted to a regular raster.
- Natural Neighbor—Finds the closest subset of input samples to a query point and applies weights to them based on proportionate areas to interpolate a value.
- Inverse Distance Weighting—Determines cell values using a linearly weighted combination of a set of sample points or cells. The weight is a function of the inverse of the distance from the known points or cells.
Cell Size—The cell size for the output raster will be automatically detected; however, you can change this if you want. The cell size can be changed, but the extent of the raster dataset will remain the same.
Search Radius—Identifies the number of cells to be included for the selected resampling method.