TC Rain Rate Azimuthal Average Documentation

Last update: 5/27/2022

This POD calculates and plots azimuthal averages for tropical cyclone (TC) rain rates from TC track data and model output precipitation.

Version & Contact info

  • Version/revision information: version 1 (5/06/2020)

  • PI Daehyun Kim, University of Washington,

  • Developer/point of contact Nelly Emlaw, University of Washington,


In the current draft code, the data is loaded in first. The netcdf file for the total precipitation is loaded in through xarray. The TC track data is read in line by line from a txt file.

Then the azimuthal average is calulated for a set of discrete radii by measuring the distance between a data point and the center of the storm for each snapshot.

A plot showing TC rain rate as a function of radius is made for select snapshopts that reach 35-45 knot max wind speeds.

Required programming language and libraries

matplotlib, numpy, netcdf, xarray, scipy

Required model output variables

Total Precipitation

TC Track Data (required: storm center latitude and longitufe, time for each snapshot, optional: max windspeed, minimum central surface pressure, sst, etc)


1. Kim, D., Y. Moon, S. Camargo, A. Sobel, A. Wing, H. Murakami, G. Vecchi, M. Zhao, and E. Page, 2018: Process-oriented diagnosis of tropical cyclones in high-resolution climate models. J. Climate, 31, 1685–1702,

2. Moon, Y., D. Kim, S. Camargo, A. Wing, A. Sobel, H. Murakami, K. Reed, G. Vecchi, M. Wehner, C. Zarzycki, and M. Zhao, 2020: Azimuthally averaged wind and thermodynamic structures of tropical cyclones in global climate models and their sensitivity to horizontal resolution. J. Climate, 33, 1575–1595,

3. Moon, Y., D. Kim, A. A. Wing, S. J. Camargo, M. Zhao, L. R. Leung, M. J. Roberts, D.-H. Cha, and J. Moon: An evaluation of global climate model-simulated tropical cyclone rainfall structures in the HighResMIP against the satellite observations, J. Climate, Accepted.

More about this diagnostic

This POD calculates the azimuthally averaged precipitation rate of tropical cyclones (TCs) as a function of distance from the center of a TC from hourly (1,3,6, or 12 hr) model precipitation output and model TC track output.

In its current version the POD requires TC track input and cannot track TCs from model output alone. The track data must provide the latitude and longitude of the storm center and the time of track snapshot and optionally a characteristic of the storm at that snapshot to use as a threshold to use the storm in the plotted average (here max wind is used and the default threshold is set to 35-45 knots to capture weak storms which meet the requirements to be considered a tropical cyclone - this threshold can be changed in the setting.jsonc file in the POD directory).

The POD takes track data separated by basin. The basin longitude regions are determined by the ECMWF TC tracking standards and are as follows: Atlantic Ocean (atl) : 100 - 350 W Eastern Central Pacific (enp) : 0 - 100 W Western North Pacific (wnp) : 100 - 180 E Indian Ocean (nin : 40 - 100 E South Indian Ocean (sin) : 30 - 90 E Australian Ocean (aus) : 90 - 160 East South Pacific Ocean (spc) : 160 E - 120 W The latitude range for all basins is 0 - 30 N or S depending on the hemispehre.

The output of this POD will be a plot of the azimuthally averaged rain rate (vertical axis) as a fuction of distance from the center of the storm (horizontal axis). The result should always be the highest rain rates at or just off center of the storm. Rain rate distributions with greatest inner-core rainfall tend to simulate stronger TCs more often. The distribution will vary greatly depending on model characteristics and threshold to determine snapshops used in average. For example was speculated in Moon et. al, 2020 that low resolution models require more energy to sustain themselves through induced precipitation.