Replacing KNMIs Present Weather Sensors: insights in systematic sensor renewal

The Royal Netherlands Meteorological Institute (KNMI) operates the national surface observation network for weather, climate, and aviation applications. Ensuring long term consistency and traceability of these observations requires systematic lifecycle management of the underlying instrumentation. This presentation provides insight into the scientific and methodological aspects of such a transition, using the replacement of KNMI’s present weather and visibility sensors as a case study.

Since approximately 2000, visibility and present weather within the KNMI network have been measured using the Vaisala FD12P Present Weather Sensor (PWS). As this instrument has reached end of life and is no longer supported by the manufacturer, KNMI initiated a replacement process that resulted in the selection of the Vaisala FD70 PWS.

To evaluate the performance of the new sensor and the impact of this transition on the observations, an extensive parallel measurement program was established. From April 2024 to January 2026, FD70 sensors were deployed at multiple stations representing a range of environmental conditions. Their performance was evaluated through comparisons with co-located FD12P instruments, and at the test field in De Bilt: reference transmissometer measurements. In De Bilt, two FD70s were also placed side to side to check for consistency and natural variability.

The analysis reveals good agreement between the new FD70 present weather sensor and the transmissometer reference for the lower visibility range (up to ~2 km). However, systematic differences in visibility were observed between the FD12P and FD70, particularly at high visibility ranges. To maintain the long term records, these differences need to investigated and (if needed) corrected for. We found that differences could be attributed to changes in instrument calibration and deterioration, sensor contamination, and insect interferences. Local variations in visibility, caused by patchy fog, introduced additional differences between the two sensors. After extensive automatic and manual filtering, FD70 data can be converted to the FD12P reference (and vice versa), enabling consistent long‑term records.

Present weather identification also shows clear improvement. The FD70’s increased sensitivity compared to the FD12P leads to more reported precipitation events and better detection of small hydrometeors, enabling more reliable distinctions between phenomena such as fog and drizzle. These findings align with independent observations from rain gauges and the KNMI validation team.

These findings highlight the challenges associated with maintaining homogeneous long term visibility and present weather records during sensor transitions. The implications extend to, for example, (aviation) meteorology, fog climatology, and air quality studies.