READ BEFORE THE TABLES

What this data is, and what it is not

The figures are 1991–2020 WMO Climatological Standard Normals, submitted by the China Meteorological Administration under the WMO programme and archived by NOAA NCEI. Five stations report 30 years used for every displayed parameter. Jinghe station 57131, mapped here to Xi'an, reports 16 years used for all nine displayed parameters.

The normal-period label and the number of contributing years are separate; neither is a forecast. A station is not the whole city: Xi'an uses Jinghe station 57131, Chengdu uses Wenjiang station 56187 and Shanghai uses Baoshan station 58362. Check a live forecast close to departure. Open the National Meteorological Centre warnings (中央气象台预警), choose your destination and check the warning's validity period; a climate normal does not clear today's activity.

July: everyone is hot; precipitation is the variable

Rain days mean days with at least 1 mm of precipitation. The table is sorted by precipitation total.

City / stationMean maxMean minPrecipitationRain days ≥1 mm
Guilin32.8°C25.2°C266.3 mm13.1
Guangzhou33.3°C25.8°C242.6 mm14.2
Chengdu — Wenjiang station 5618730.0°C22.1°C231.8 mm12.1
Beijing31.8°C23.0°C169.7 mm8.8
Shanghai — Baoshan station 5836232.4°C25.9°C146.9 mm8.5
Xi'an — Jinghe station 57131 (16 years used)32.8°C23.4°C84.9 mm7.3

The mean-maximum spread is 3.3°C. The precipitation-total spread is 3.1×. Guilin's monthly total is more than three times Xi'an's Jinghe total.

January: a 17.5-degree country

City / stationMean maxMean min
Guangzhou18.7°C10.6°C
Guilin11.7°C6.1°C
Chengdu — Wenjiang station 561879.3°C2.7°C
Shanghai — Baoshan station 583628.2°C2.2°C
Xi'an — Jinghe station 57131 (16 years used)4.9°C−3.1°C
Beijing2.3°C−6.9°C

The gap between Beijing's and Guangzhou's January mean minimum is 17.5°C. A route covering both ends is two climates in one suitcase.

October: a cooler alternative to July

These are whole-October normals, not mid-month or late-month figures. Beijing, Jinghe and Baoshan are cooler and have lower precipitation totals than July, while their mean maxima and minima are higher than January. This supports an October comparison; it does not guarantee weather on chosen dates or prove that late October is best.

City / stationMean maxMean minPrecipitationRain days ≥1 mmYears used
Beijing — station 5451119.3°C8.8°C28.7 mm3.330
Xi’an — Jinghe station 5713119.7°C11.5°C52.5 mm6.816
Shanghai — Baoshan station 5836222.9°C16.7°C71.4 mm5.130
Chengdu — Wenjiang station 5618720.7°C14.7°C42.6 mm8.430
Guilin — station 5795726.3°C18.3°C54.9 mm5.330
Guangzhou — station 5928729.3°C20.9°C67.3 mm3.830

The shape of the year, by station

City / stationAnnual precipitationRain days ≥1 mmPlanning interpretation
Guilin1,988.0 mm121.6Highest total in this set and frequent wet days
Guangzhou1,950.1 mm109.8High totals and the most July rain days in this comparison
Shanghai — Baoshan station 583621,236.0 mm94.6Mid-table total, with precipitation spread across many days
Chengdu — Wenjiang station 56187909.3 mm90.7Lower total than Shanghai but a similar count of wet days
Xi'an — Jinghe station 57131 (16 years used)542.1 mm57.9Low annual total and the lowest July total in this set
Beijing528.0 mm43.7Lowest annual total, but July alone contributes about 32%

Beijing is dry annually but concentrated in summer. An annual total can hide the month that controls your route.

Winter light is a separate decision

January sunshine changes outdoor photography and long walking days even where temperature looks tolerable.

01

Beijing

188.1 hours

02

Guangzhou

112.9 hours

03

Shanghai — Baoshan

110.4 hours

04

Xi'an — Jinghe (16 years used)

109.7 hours

05

Guilin

58.4 hours

06

Chengdu — Wenjiang

45.4 hours

Six worked cases

M1
DEFAULT

October, multi-city

Conditions
A Beijing, Xi’an and Shanghai route when the month is flexible.
Result
Compare the October rows above with January and July. For Beijing, Jinghe and Baoshan, October has lower mean temperatures and precipitation totals than July and warmer mean temperatures than January. Choosing dates after the National Day holiday is a separate calendar decision; the monthly climate normals cannot distinguish early, middle and late October. Use the holiday calendar before booking and an actual forecast near departure.
M2
FAILURE

July outdoor route through Guangzhou and Guilin

Conditions
The route depends on exposed walking days.
Result
Guangzhou has 14.2 and Guilin 13.1 July rain days of at least 1 mm. Move the month, reduce outdoor dependency or pair an indoor alternative with every exposed day.
M3
REVERSAL

“Beijing is dry, so July is safe”

Conditions
The annual total is mistaken for the selected month.
Result
Beijing has the lowest annual precipitation here, yet July contributes about 32% of it. Dry annually does not mean dry in July.
M4
FAILURE

One clothing plan for Beijing and Guangzhou in January

Conditions
One suitcase must cover both ends.
Result
Their January mean minimums are 17.5°C apart. Use layers and separate cold-weather gear, split the route by region or move to a shoulder season.
M5
FAILURE

Treating a normal as a booking-date forecast

Conditions
The table's 32.8°C Jinghe value is used as a date prediction.
Result
It is the 1991–2020 mean daily maximum at one station, not a promise, ceiling or forecast for your dates.
M6
EDGE

Calling Jinghe and Wenjiang city-centre weather

Conditions
The editorial city mapping is mistaken for a source-station name.
Result
This page maps Jinghe 57131 to Xi'an and Wenjiang 56187 to Chengdu while keeping the real stations visible. Do not use either as block-level conditions for a hotel.

Failure recovery

R1

Your dates are fixed in a wet month

Trigger
A fixed date lands in the wetter part of the route.
First action
Put the least weather-sensitive city or indoor-heavy days into that period.
Alternative
Change the city order rather than cancelling the trip.
Escalation
If the outdoor activity has no acceptable indoor or route fallback, consider changing the month. No month guarantees usable weather; the forecast and official warning still decide the day.
R2

Your winter route spans north and south

Trigger
One itinerary crosses the 17.5°C January minimum gap.
First action
Pack layers and one genuinely warm outer item.
Alternative
Leave the heavy item at a hotel during the southern leg or split the trip by region.
Escalation
None; this is a routing and luggage decision.
R3

You need actual conditions

Trigger
A booking decision needs the weather on exact dates.
First action
Use a live official forecast close to departure. Check the official warning for your destination and its validity period; a climate normal does not clear today’s activity.
Alternative
Use this page only to compare cities and months.
Escalation
For hazardous weather, follow official warnings rather than a climate table.
R4

You need an exact neighbourhood

Trigger
A hotel, mountain or district microclimate controls the decision.
First action
Note which station represents the city and its relationship to your destination.
Alternative
Use a closer official station carrying the same normal period where available.
Escalation
Do not relabel or blend stations without documenting the mapping.

What this page does not cover

  • A forecast for your dates or conditions outside these six stations
  • Neighbourhood microclimates or block-level hotel conditions
  • Typhoons, heat waves, sandstorms or other episodic events
  • Holiday crowding, prices or attraction opening hours
  • A guarantee that spring or autumn will be comfortable in a particular year
  • Any replacement of live official warnings close to departure
DO THIS NOW

Use the comparison in five steps

  • Find your cities in the January, July and October tables.
  • Compare rain days and precipitation before small summer temperature differences.
  • Treat a north–south winter route as two climates.
  • Check the 2026 holiday calendar before booking.
  • Use a live official forecast and destination warnings close to departure; check the validity period before any outdoor block.

Sources and verification boundary

The station files were fetched from the NCEI WMO archive on 18 August 2026 and reproduced on 20 September 2026. CMA submitted the data under the WMO programme; it was not downloaded from a CMA webpage. The six original CSVs and checksums are retained in the source archive. Current remote CSV downloads timed out in the latest check, so this is a reproduction of the retained historical files, not a fresh download. Xi'an–Jinghe and Chengdu–Wenjiang remain editorial mappings. The October table uses the same source rows and period as January and July.

WMO Climatological Standard Normals product pageNOAA National Centers for Environmental Information ↗NOAA release note: 1991–2020 WMO normals and station CSV filesNOAA National Centers for Environmental Information ↗