Temperature Inversions in the Bergen Valley and Testing of Low-Cost Temperature and Air Quality Measuring Instruments

Abstract

This study aimed at two goals, the testing of simple low-cost measurement sensors for temperature and air pollution, and the investigation of temperature inversions in the Bergen valley. For this we conducted a 13 months measurement campaign between February 2017 and February 2018. In addition to already existing measurement stations we deployed a dense vertical transect of stations along a mountain side directly next to the city center. The largest part of the ground-based temperature inversions were found to be shallow, most frequently with tops not exceeding 100 m asl. These were in general short lasting with more than 75 % shorter than two hours. Ground-based inversions are associated with high pollutant concentration events, but during the campaign no such events occurred. Elevated inversions were found to occur most frequently between 225 and 300 m asl and again more than 75 % lasted less than two hours. The synoptic situation during the study period, with a lot of frontal passages and a total amount of precipitation of more than 3000 mm, was likely to be the reason for the low number of long lasting inversions. The dense network of stations was instead used for a detailed study of a strong but comparatively short lasting temperature inversion over the Bergen valley caused by warm air advection. The testing of low-cost temperature loggers (EasyLog), showed that solar radiation was a major problem. The loggers were prone to overheating when exposed to solar radiation. Removing all data when the solar radiation exceeded 20 W/m^2, resulted in a RMSD that decreased from a value between 1.7 and 3.6 % to a value between 0.4 and 0.7 %, and a increase in the correlation from a value of 0.58 and 0.82 % to 0.98 %. Testing of low-cost air quality measuring instruments (Air Quality Egg), showed a questionable and non-reproduceable behavior. When comparing the Air Quality Egg to an official air quality measuring station in Bergen, we found a RMSD of 128.9 % and a correlation of 0.24, for the Air Quality Egg measuring NO2. After intensive testing of these instruments we therefore decided to abandon the initial plan of field deployment of the Air Quality Eggs and advice against their usage for scientific studies without further thorough testing.