Performance and longevity of compact all-in-one weather stations – the good, the bad and the ugly

Abstract

Abstract. We provide a long-term evaluation of compact, all-in-one automatic weather stations (AiOWS) compared to professional-grade Automatic Weather Stations (AWS). We examine the performance, longevity, and degradation of six AiOWS models over several years of non-serviced use. The objective was to determine how closely these low-cost stations meet World Meteorological Organization (WMO) performance standards for temperature, humidity, wind, and precipitation, and to identify their weaknesses and maintenance needs. Previous studies show the potential value of AiOWS when data are properly quality-controlled, yet long-term reliability remains uncertain. To address this we deployed six AiOWS units – Davis VVue, Davis VP2, METER ATMOS41, Lufft WS601, and Vaisala WXT520, alongside two collocated reference AWS meeting WMO standards. Before field installation, each unit was tested in the calibration lab of the Royal Netherlands Meteorological Institute (KNMI) for baseline validation. The stations were then operated in open terrain for multiple years without any servicing, simulating worst-case-scenario end-user neglect. Initially, all AiOWS met manufacturer specifications. After long-term exposure, however, sensors displayed varied durability. The Vaisala unit operated continuously for over 13 years, while others failed between 4 and 10 years due to corrosion, component wear, and sensor drift. The METER and Davis VVue remained mostly functional but with degraded performance, whereas both Davis VP2 rain gauges failed early due to reed switch damage. Temperature measurements were the most robust. In climate chamber tests, new and aged sensors maintained accuracy within ± 0.3 °C across −15 °C to 30 °C, drifting slightly (underestimating by 0.5–0.7 °C) above 30 °C. Field data confirmed these results, though strong solar radiation caused overestimations during summer. The Vaisala and Davis VVue units remained within WMO Class B limits after multiple years. Relative humidity showed consistent deterioration. Most sensors overestimated low humidity and underestimated above 90 %, particularly the METER unit, whose bias grew markedly after 5 years. Wind speed accuracy degraded due to mechanical wear. Cup anemometers underreported low winds and failed completely in some cases. Sonic sensors (Vaisala, METER) produced erratic readings after several years, highlighting their fragility outdoors. Precipitation performance was weakest across all models. Tipping bucket designs suffered from clogging, internal corrosion, and undercatch errors, while haptic or drip-based sensors became inaccurate as components aged or fouled. We concluded that compact AiOWS can provide scientifically useful temperature data if properly managed but fall short for humidity, wind, and particularly precipitation unless regularly serviced. Long-term unattended operation severely limits reliability, yet moderate maintenance can potentially restore performance close to WMO Class A/B standards, extending their utility for dense observation networks.