python-TAMER 0.4.0

Toolkit for Analysis and Maps of Exposure Risk

Introduction to python-TAMER

Official documentation hosted at https://tch521.github.io/python-TAMER

Python-TAMER is a python package for performing advanced environmental exposure calculations in a simple and accessable way suitable for users with minimal python experience. To get started, check the Installation tab.

• Free software: BSD-3-Clause licence

The Toolkit for Analysis and Maps of Exposure Risk (TAMER) is a python package designed to calculate estimates of individual and population exposure across a geographic area. Currently, the project is focused on erythemal UV radiation exposure, but the tools provided by TAMER could be used in a variety of contexts provided there is appropriate source data. In addition to providing a simple methods for basic exposure calculations, (such as mean, median, and maximum intensities over certain time periods,) TAMER allows users to calculate daily doses by integrating the exposure over time. TAMER deals with very large volumes of data, but is designed with memory efficiency in mind so that such data can be processed on even modest personal comptuters.

In the context of UV, the dose received by an exposed individual is far more relevant to their corresponding health risk than the ambient level of UV. Usually, these doses are measured by wearable devices. Harris et al. 2021 (https://doi.org/10.3390/atmos12020268) explains the various benefits of instead estimating such doses based on satellite-derived data with sufficiently high spatial and temporal resolution. The Swiss UV climatology provided by Vuilleumier et al. 2020 (https://doi.org/10.1016/j.envint.2020.106177) is Currently the most appropriate source data for TAMER as it provides erythemal UV at an approximately 1.5km spatial resolution and an hourly temporal resolution. Harris et al. 2021 shows that with location, date, and time information, reasonable ambient doses can be calculated. However, to calculate personal doses, the Exposure Ratio (ER) must also be known, that being the ratio between the ambient dose and the personal dose received by a certain body part. Different body parts have varying ERs which also depend on body posture, for example the ER of the forehead is lower when bowing down than when standing normally. TAMER includes a model from Vernez et al. 2015 (https://doi.org/10.1038/jes.2014.6) to calculate ERs according to anatomic zone, posture, and time of year. python-tamer.SpecificDoses is a table-like class designed to take location, date, time, posture, and anatomic zone information to calculate specific ambient and personal doses of the described individuals.

A large part of TAMER is dedicated to producing high quality maps of a variety of exposure metrics. Maps of UV exposure often show the mean, median, or max irradiance for a given time period. TAMER includes the option to calculate such maps, but also offers more advanced alternatives. The TAMER approach balances versatility with memory efficiency by calculating histograms for each pixel as a first step. These histograms can describe the irradiance or the daily doses for any time selection and exposure schedule. They can be built up iteratively, processing one year at a time to ensure only moderate memory usage. With the pixel histograms calculated, the user then has to choose a statistical descriptor to condense the distribution into a single number to be plotted on the map. This can be basic statistics such as mean, median, or max, however we include some more advanced options such as a custom percentile and the standard deviation. In a forthcoming release, we shall also include the option to define one’s own formula for a custom descriptor, allowing for metrics like the difference between the 95th percentile and the median divided by the standard deviation which would be indicative of the severity of acute exposure instances. The simple and novel approaches to exposure estimation provided by the combined release of high resolution UV data (https://doi.org/10.1016/j.envint.2020.106177) and the simple and novel exposure calculations provided by TAMER give opportunity to epidemiologists and public health experts to study UV exposure with higher detail than has ever been possible before.

Features

• Calculate daily doses rapidly with custom exposure schedules

• Analyse exposure distributions per pixel

• Produce maps to represent chronic and acute exposure using standard or custom metrics

• Replicate dosimetry measurements using Exposure Ratio modelling

In Development

• Improved support for custom statistical descriptors

• Custom area selection for the SpecificDoses class

Future work

• Improved support for different source files (new units, temporal resolutions, etc.)

• Integrate support for cross multiplication of ExposureMap with population distribution data

History

0.4.0 Open Alpha (2021-05-24)

• Added the new ExposureMapSequence class providing much improved map making capabilities (this was the majority of the work)

• Added the render_map() function to produce more beautiful maps across in a more versatile way

• Added the str2daysofyear() function to interpret keywords like month names and produce the days of the year

• Added the analyse_variable() function to the SpecificDoses class for more flexible analysis options

0.3.1 Open Alpha (2021-03-29)

• Fixed errors in example code in documentation

• Fixed ER_Vernez_2015() function

• Updated assumptions about units to reflect new dataset from Vuilleumier in UVI rather than W m-2

• Fixed issue where modifying map_options property would result in it being set to None

0.3.0 Open Alpha (2021-03-23)

• Compiled and added to PyPI for easy public access

• Added standalone function for calculating ER using the Vernez 2015 method: ER_Vernez_2015()

• Significantly expanded and standardised docstrings, adding examples

• Fixed error involving day selection in SpecificDoses class being one day late

• Added SpecificDoses.standard_column_names() function for standardising column names to ensure functionality

0.2.0 Alpha (2021-03-11)

• Added documentation

• Added basic unit tests for each class (SpecificDoses and ExposureMap)

• Added histogram descriptor calculator functions to subroutines.

• Added map making functionality for ExposureMap class, limited consideration of map_options at this stage

• Fixed errors when working with single day test data (but anticipate further issues with this, to be fixed in a later release)

0.1.0 Pre-Alpha (2021-03-02)

• Alpha release on github only, no documentation and limited functionality