An algorithm of area filling based on isoline classification is presented. Based on the relationship between isolines and borderlines, isolines are classified and ranked, and the isoline areas are constructed. The isoline area topology is built by searching the subisolines. The advantages and disadvantages of different ways of color filling are also discussed, and the analysis concludes that it is necessary to use grid data to choose color. This algorithm can be used to fill the areas between regular isolines for less computing time. The algorithm has been applied successfully in the Shanghai Nowcasting & Warning System (NoCAWS).