Premium
A Physical Antialias Filter for Time‐Series Temperature Measurements
Author(s) -
Fairley Jerry P.,
Zakrajsek John R.
Publication year - 2007
Publication title -
groundwater monitoring and remediation
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.677
H-Index - 47
eISSN - 1745-6592
pISSN - 1069-3629
DOI - 10.1111/j.1745-6592.2006.00128.x
Subject(s) - nyquist frequency , nyquist–shannon sampling theorem , impulse invariance , anti aliasing filter , filter (signal processing) , thermistor , signal (programming language) , sampling (signal processing) , frequency response , electronic engineering , analogue filter , computer science , thermocouple , oversampling , signal processing , low pass filter , acoustics , engineering , high pass filter , telecommunications , electrical engineering , digital filter , digital signal processing , bandwidth (computing) , physics , infinite impulse response , programming language
Collecting high‐quality time‐series data requires sampling at a frequency at least twice that of the highest frequency in the monitored signal (the Nyquist frequency). Many signals of interest to hydrologists vary slowly over long periods of time but display infrequent and/or unpredictable episodes of rapid, high‐frequency change that can result in a corrupted (aliased) data stream. Although the standard remedy for this difficulty is to use an analog filter to remove frequencies above the Nyquist frequency prior to sampling, electronic components to construct analog filters for frequencies of interest to hydrologists do not exist. To mitigate this problem when monitoring temperatures in natural systems, we suggest enclosing the temperature monitoring device (thermocouple or thermistor) in a housing designed to change the characteristic response time of the device and thus create a band‐limited signal suitable for sampling. We call this strategy “physical antialias filtering” because it depends on manipulating physical heat transfer characteristics rather than analog electronics to filter out frequencies above the Nyquist. We provide a simple analytical expression to calculate the frequency response of a filter with a known characteristic response time and illustrate its application with the results of laboratory tests. In addition to its usefulness in designing antialias filters, the analytical expression provided can be applied to the correction of data from a temperature probe that is housed to protect it from contact with corrosive fluids or other damaging environmental influences.