A note on downward continuation of gravity

Abstract Comparing calculated and exact values in an ideal example, it is shown that Peters' proposed method of downward continuation of gravity is far less suitable than a combination of Peters' procedure for upward continuation and a finite differences method for downward continuation. It is shown also how Peters' method can be substantially improved by considering only the closest values in the downward step of a continuation process.