Continuous observations of the atmosphere and clouds are essential for understanding snowfall processes. While ground-based microwave radiometers (MWRs) are effective tools, their accuracy degrades when liquid water or snow accumulates on the radome. Standard built-in blowers often fail to prevent snow accumulation in heavy snowfall regions. To address this, we implemented a high-capacity external blower as a snow accumulation countermeasure. Precipitable water vapor (PWV) was observed in Nagaoka, Niigata, from November 2024 to March 2025 using two identical compact MWRs: one with the external blower (BL) and one without it (NOBL). PWV was retrieved using a machine-learning model and evaluated against the Japan Meteorological Agency Local Analysis (LA) data. Under snow-free conditions, both units exhibited high consistency (mean error [ME] = 0.14 mm, root mean square error [RMSE] = 1.04 mm), indicating that both can be utilized to verify the effectiveness of snow accumulation countermeasures. During snowfall events, the accuracy of the NOBL unit significantly decreased (RMSE = 2.51 mm) due to snow accumulation and subsequent meltwater film formation on the radome. In contrast, the BL unit maintained accuracy comparable to non-snow conditions (RMSE = 1.68 mm). Furthermore, data availability during snowfall reached 99.9% for the BL unit, compared to only 67.8% for the NOBL unit. These results demonstrate that the external blower system is highly effective for maintaining the reliability and continuity of MWR observations in heavy snowfall areas, marking a significant technical advancement for establishing observation networks in such challenging environments.

A short-term extreme heavy snowfall occurred over the Tokachi Plain in Hokkaido on 3–4 February 2025. Because a marine heatwave (MHW) was identified offshore during the event, this study investigates the characteristics of heavy snowfall in Tokachi and the uniqueness of this case through statistical analyses focusing on atmospheric and oceanic conditions. The preceding 1-h, 3-h, 6-h, and 12-h snowfalls were the largest on record and deviated substantially from regression relationships, indicating an anomalous event. An analysis of snowfall cases of 1953–2025 revealed three synoptic patterns: WSCC, including the 2025 case, characterized by cyclones along both the western and southern coasts of Hokkaido (44%); SCC, with a cyclone only along the southern coast (41%); and WCC, with a cyclone only along the western coast (13%). Surface observations from 1998 to 2025 show that a cold-air layer formed by the geographical characteristics of the Tokachi Plain maintained low temperatures while supporting abundant precipitation, leading to short-term heavy snowfall. A cold-air layer was present in all three patterns, and a coastal front was suggested in most cases. This case was distinguished by remarkably cold upper-level air combined with lower-to-middle-tropospheric warming and moistening upstream due to the MHW. WSCC cases were subdivided into two groups depending on which cyclone, either western or southern, was more developed, resulting in differing southern cyclone positions and snowfall mechanisms. Remarkably heavy snowfall occurred only in the group with a more developed western cyclone, including this case, which also exhibited the most unstable atmospheric conditions among these cases.