Local tallies from Wenzhou, Shanghai and Fujian showed more than 1.2 million relocations by Sunday evening after Typhoon Dolphin came ashore twice in Zhejiang and sent torrential rain across eastern China.

China’s National Meteorological Center said Dolphin made its first landfall at about 5:30 p.m. on August 9 near Kanmen in Yuhuan, part of Zhejiang’s Taizhou city, with maximum winds of 42 metres per second, or about 151 kilometres per hour. The centre crossed the coast again about 70 minutes later near Yueqing, south of the first landfall. By early Monday it had weakened to a tropical storm, but that change in classification did not end the emergency: the storm’s broad rain shield continued moving northwest.

The evacuation count grew sharply as local authorities expanded operations and released later figures. Wenzhou, a coastal city of nearly 10 million people, reported relocating more than 900,000 residents and opening over 1,500 shelters. Shanghai said 215,600 people had been moved from dangerous areas, while Fujian reported another 98,900 relocations. Those three geographically separate tallies alone exceed 1.2 million.

An Associated Press report issued before landfall contained an internal mismatch: its lead said more than 300,000 people had been relocated, while its named figures included about 390,000 in Taizhou and roughly 30,300 in Shanghai. Later reports supplied the larger Wenzhou, Shanghai and Fujian numbers. Adding the separately reported Taizhou figure would take the sum past 1.6 million, but China had not released a consolidated national total and the local updates came at different times. The headline therefore describes a conservative sum of non-overlapping local reports, not an official national tally.

Shanghai endured extraordinary rainfall even though Dolphin’s centre crossed the coast well to the south. The Xujiahui weather station in central Shanghai measured 312.9 millimetres in the 24 hours ending at 7 a.m. Monday, the highest 24-hour total in records dating to 1872. A station in Tangqiao, Pudong, approached 400 millimetres. Streets became channels and water entered some neighbourhoods as the storm’s spiral bands repeatedly crossed the city.

An official Shanghai report said monitoring systems identified more than 180 waterlogged locations in Pudong, Xuhui, Huangpu and Changning districts. By late Sunday afternoon crews had cleared 155 of them, while more than 300 flood-control pumping stations operated at full capacity and discharged 55.71 million cubic metres of water. In Yueqing, state broadcaster CCTV reported searches for people trapped in flooded homes and crews clearing fallen branches. Authorities continued to warn of flash floods and landslides as rain spread inland.

Cited Monday updates reported no deaths in mainland China. In the Philippines, a landslide during monsoon rains intensified by Dolphin killed five people; AP put the broader weeklong northern Luzon weather toll at 15, not all attributable to the typhoon. The distinction matters because a regional disaster total should not be presented as a toll caused by one storm.

Transport disruption was extensive. Shanghai’s Hongqiao and Pudong airports cancelled more than 1,300 flights on Sunday, according to AP and the Guardian, while local rail and metro services were suspended or curtailed. Zhejiang halted ferries and cruises before landfall. Hangzhou began restoring subway service Monday morning, and Ningbo airport planned to resume flights at midday while warning that delays and cancellations could continue. The uneven restart reflected a difficult trade-off: every hour of closure carries an economic cost, but reopening before tracks, bridges and runways are inspected can compound the danger.

By Monday evening, the National Meteorological Center placed Dolphin’s weakened centre in Jiangxi and forecast that it would continue northwest before dissipating. The agency still expected heavy rain across Zhejiang, Shanghai, Jiangsu, Anhui, Henan, Hubei, Hebei and Shandong, with locally extreme totals possible in central China. Wind strength was falling, but saturated slopes, swollen waterways and urban drainage systems remained the more persistent threat.

The scale of the relocations demonstrates real administrative and logistical capacity, and moving people early is one of the strongest tools available to prevent deaths. But a dramatic evacuation number is not, by itself, a complete measure of success. Shelter conditions, access for elderly and disabled residents, accurate casualty reporting and safe return plans matter after the television images fade. So does a clear account of which residents were moved, when and from where.

That data gap is especially important in China, where official information is tightly managed and authorities have recently intensified action against alleged online disaster misinformation. False videos and invented death tolls can obstruct rescues, but public trust also depends on the government releasing prompt, detailed and internally consistent facts. Shanghai’s disclosure of waterlogging points, pumping capacity and relocation totals shows what useful reporting can look like; a timely national ledger would make it easier to test claims, direct aid and judge performance.

Dolphin also exposed the limits of infrastructure designed around older rainfall expectations. Shanghai’s pumps moved an enormous volume of water, yet record rain still overwhelmed streets. A dedicated attribution study would be needed to measure how climate change affected this storm. The planning problem is broader: warmer air can carry more water, making yesterday’s rainfall records a weaker basis for tomorrow’s drainage standards. Eastern China’s ports, factories and megacities need transparent risk maps, upgraded systems tested against new extremes, and public accountability for whether promised protections actually work.