Network requirements for sensor accuracy and precision: a case study to assess atmospheric variability in simple terrain

摘要

This simulation-based study is an effort to develop design requirements for a hypothetical network of instruments to measure average atmospheric variability over a relatively simple terrain. A dataset of air and soil temperature, relative humidity, wind speed, and soil moisture from the vicinity of Lincoln, Nebraska, was used to simulate atmospheric daily measurements (max and min temperature as well as infra-red surface temperature (grass cover), relative humidity, wind speed and soil moisture) at various distances from the site. An analysis was performed to determine the instrumentation requirements needed to distinguish natural atmospheric variability from input instrumental error at the 0.05 significance level. Both accuracy and precision requirements were examined to determine the affect of systematic and random errors, respectively. It was found that sensor accuracy (small systematic bias relates to high accuracy) does not affect the ability to determine the atmospheric spatial variability. However, sensor precision (small random error equates to high precision) was found to affect the ability to determine the spatial variability of the atmospheric variable in question; commercial sensors are currently available that meet or exceed the precision requirements for all except one of the variables examined. Copyright (c) 2007 Royal Meteorological Society.