Specifies which mathematical algorithm is to be used to analyze the data field.

• Nearest neighbor — This is by far the quickest, and most users will use this setting. Values are mapped to the closest gridpoint, then are mathematically expanded to neighboring unfilled gridpoints in a cyclical manner. Afterwards, a selective smoothing operator smooths gridpoints, performing more smoothing iterations performed on gridpoints that are further from representative data areas and less or no smoothing passes where data is close to a gridpoint. Strengths: It's fast and always works. Weaknesses: Data void border areas look linear, and data void interior areas can look pie-shaped.
• Weighted — Digital Atmosphere looks at the neighboring stations around a gridpoint, giving it an average of those values around it with more emphasis on the values at closer stations.
• Barnes — The Barnes method, the cornerstone of the MCIDAS and GEMPAK weather workstations, has a number of advantages: versatility, simplicity, and speed. Each gridpoint is assigned a meteorological value which is the result of a Gaussian distance-dependent weighting function. A difference field is calculated, which determines how far off from reality the analysis field appears to be, then a difference correction is applied to the analysis field. This takes a relatively long time to compute, but is one of the more accurate analysis algorithms.
• Cressman — Each gridpoint is assigned a meteorological value which is the result of a distance-dependent weighting function. The Cressman analysis method is similar to the Barnes technique, however it makes successive corrections using a smaller and smaller radius of influence to eliminate all errors. It takes more time, but the results are good. In data-sparse areas, it is common for there to be erroneous contours, especially if Digital Atmosphere is tasked to determine the best station spacing. Strengths: It's sophisticated and flexible, and it may retain more mesoscale structure. Weaknesses: It can't resolve short-wavelength components, it can require a large amount of time, and it can fail with extreme data distribution.

ANALYSIS SETTINGS 

These are used to control the performance of the selected analysis type.  Certain choices may be grayed out depending on the analysis type selected and whether "automatic smoothing" is in effect.