Tropical Cyclone Center-Fixing in Microwave or Infrared Imagery

Precise center-fixing of tropical cyclones (TCs) is critical for operational forecasting, intensity estimation and visualization. Current procedures are usually performed with manual input from a human analyst, using multispectral satellite imagery as the primary tools. While adequate in many cases, subjective interpretation can often lead to variance in the estimated center positions. In this paper we present an objective, robust algorithm for resolving the rotational center of TCs: the Automated Rotational Center Hurricane Eye Retrieval (ARCHER). The algorithm finds the center of rotation using spirally-oriented brightness temperature gradients in the TC banding patterns in combination with gradients along the ring-shaped edge of a possible eye. It is calibrated and validated using 85-92 GHz passive microwave imagery because of this frequency's relative ubiquity in TC applications, however, similar versions of ARCHER are also shown to work effectively with 37 GHz and infrared imagery of TCs. In TC cases with estimated low to moderate vertical wind shear, the mean accuracy of the ARCHER estimated center positions is 17 km (9 km for Category 1-5 hurricanes). In cases with estimated high vertical shear, the accuracy of ARCHER is 31 km (21 km for Category 2-5 hurricanes).