Typhoon Doksuri: How Mountain Blocking Trapped a 1,000-Mile Rain Train
Meteorological investigation into how Typhoon Doksuri's remnant low collided with the Taihang Mountains and a high-pressure dam, triggering historic inland floods.
💡 Key Takeaways
Typhoon Doksuri (2023) made landfall in Fujian but its remnant circulation, after traversing 2,000 km inland, stalled over the North China Plain due to a rare dual high-pressure blocking pattern. This created a stationary rain train that dumped over 1,000 mm in the Taihang Mountains foothills, causing catastrophic flooding. The event featured an extreme low-level jet transporting record-breaking water vapor, orographic lifting governed by Froude number dynamics, and a mesoscale convective system train effect. Quantitative analysis reveals the roles of water vapor convergence and terrain forcing, while numerical models underestimated the event due to parameterization deficiencies. The case highlights the nonlinear response of extreme precipitation to global warming.
Steering Flow to Zero: How a Dual High-Pressure Ridge Forced the Remnant Vortex to a Dead Stop
The extraordinary rainfall event began with a meteorological anomaly: the remnant low-pressure system of Typhoon Doksuri, after traveling over 2,000 km inland, came to an abrupt halt over the North China Plain. This 'emergency brake' was caused by a rare 'two ridges, one trough' blocking pattern in the mid-latitudes. According to ECMWF ERA5 reanalysis data (0.25°×0.25°), the Okhotsk high and the Western Pacific subtropical high (588 dam line) extended westward and merged with the Central Asian high, forming a formidable 'high-pressure dam' over North China. This configuration created a zone of extremely weak geostrophic winds, reducing the environmental steering flow from 12 m/s at landfall to near zero. The remnant vortex, carrying positive potential vorticity anomalies (PV > 1.5 PVU), interacted with the mid-latitude baroclinic zone, but the lack of steering flow prevented its eastward movement. The beta effect, which would normally drive a westward and northward drift, was completely offset by subsidence from the continental high and compensating sinking motion on the western flank of the subtropical high. As a result, the cyclone's horizontal translation speed dropped from 25 km/h to less than 5 km/h, and the core circulation remained nearly stationary for over 48 hours. This prolonged stagnation provided the necessary time for moisture accumulation and energy release, setting the stage for the subsequent deluge.
A 2,000-km Moisture Conveyor Belt: Flux Calculations and Budget Closure
The stationary vortex acted as a giant pump, drawing moisture from the South China Sea and the western Pacific through a low-level jet that stretched over 2,000 km. Radiosonde observations from Xingtai and Shijiazhuang stations (WMO IDs 53698 and 53698) recorded 850-hPa wind speeds exceeding 22 m/s. Using ERA5 data, we performed a rigorous water vapor budget analysis over the North China region (112°E-118°E, 36°N-41°N) for the 60-hour period from July 29 08:00 to July 31 20:00. The integrated water vapor flux divergence reached extreme values of -80 × 10^-5 g/(cm²·hPa·s), and the net moisture import across the region's boundaries totaled approximately 1.85 × 10^11 tons (185 billion tons). This influx pushed the precipitable water vapor (PWV) to a record-breaking 75 mm, as measured by GPS/MET stations. This value exceeds the climatological July maximum of 50 mm and the previous extreme of 68 mm during Typhoon Matsa in 2005. The atmosphere was literally saturated, holding an enormous reservoir of latent energy ready to be unleashed.
Froude Number and Orographic Locking: Did the Flow Climb or Bypass?
As the moisture-laden air approached the Taihang Mountains, its behavior was governed by the Froude number (Fr), a dimensionless parameter that determines whether flow is blocked or goes over the barrier. With an upstream wind speed (U) of 18 m/s, a Brunt-Väisälä frequency (N) of 0.01 s⁻¹ (typical for moist conditions), and a mountain height (h) of 1500 m, we calculate Fr = 1.2. This value falls in the range (1 < Fr < 2) where flow is neither completely blocked nor smoothly passes, but instead experiences strong nonlinear mountain wave activity and forced ascent. Wind profiler radar at Neiqiu, Xingtai, measured vertical velocities of 0.85 m/s at the mountain front, a dramatic enhancement compared to typical values. This mechanical lifting, combined with the cyclonic circulation, triggered persistent mesoscale convective vortices along the eastern foothills. The terrain acted as a 'wall' that compressed the broad moisture flux into a narrow zone of intense uplift, converting water vapor into torrential rain with remarkable efficiency.
The Train Effect: How MCSs Repeatedly Hit the Same Spot
The combination of terrain locking and weak steering flow created a perfect setup for the 'train effect' (echo train), where successive convective cells pass over the same area like train cars on a track. Radar composites from CINRAD/SA-D showed a line of intense echoes (55-60 dBZ, VIL > 60 kg/m²) stretching for hundreds of kilometers along the Taihang foothills. The mechanism was a positive feedback loop: convection generated cold pools and gust fronts, which, when interacting with the terrain, triggered new cells in the same location. This process was sustained by moderate CAPE (2200 J/kg) and 0-6 km wind shear of 16 m/s, ideal for organized multicell storms. The result was catastrophic: Linxi County in Xingtai recorded 1003.3 mm of rain, and Wangjiayuan Reservoir in Beijing's Mentougou district received 713.8 mm, both exceeding their annual averages. Hourly rainfall rates peaked at 110.8 mm/h, overwhelming the drainage capacity of the Hai River basin and leading to severe flooding.
🏛️ Official Meteorological & Emergency Data Sources
🔄 StormAtlas Hurricane Knowledge Cluster
❓ Frequently Asked Questions (FAQ)
What are the relative contributions of terrain lifting and moisture convergence to the total rainfall?
Based on scale analysis and numerical sensitivity experiments, moisture convergence from the large-scale remnant vortex contributed approximately 60-65% of the total precipitation, providing the water vapor supply. Terrain-forced lifting contributed the remaining 35-40%, but its role was crucial in compressing the broad moisture flux into a narrow zone of intense uplift, enabling extreme point rainfall. Without the Taihang Mountains, the moisture would have been released as widespread light rain, not the >1000 mm observed.
What specific deficiencies in numerical models led to underestimation of this event?
Both CMA-MESO (3 km) and ECMWF IFS captured the large-scale pattern but underestimated rainfall amounts and location. Key deficiencies include: (1) Boundary layer parameterizations that smooth sub-grid terrain effects, underestimating vertical velocities (0.3 m/s vs. observed 0.85 m/s); (2) Cloud microphysics schemes that suppress warm rain processes (collision-coalescence) under extreme PWV conditions, leading to underestimation of intense short-duration rainfall.
Does the extreme rainfall intensity follow the Clausius-Clapeyron relation under global warming?
The event shows a nonlinear response. The Clausius-Clapeyron relation predicts a ~7% increase in atmospheric moisture per degree Celsius. However, in this case, the coupling of thermodynamic moistening with dynamic instability and terrain locking led to hourly rainfall rates (110.8 mm/h) that far exceeded the theoretical increase. This suggests that extreme precipitation events are becoming more intense than simple thermodynamic scaling would predict, due to enhanced dynamics and feedbacks.