How often are these flights late or cancelled?
China Southern
In the twelve months from July 2025 to June 2026, 13.2% of 1,056 movements arriving into the UK were disrupted — more than fifteen minutes late or cancelled at short notice.
Put the other way round, 86.8% of 1,056 movements met the fifteen-minute convention. These are historical observations, not a forecast.
Definitions, denominators and limits
Arrivals into UK reporting airports only unless labelled departures. Denominator: operated movements plus short-notice cancellations (M+C). A movement is not a scheduled flight. Unmatched records are data quality, not outcomes. Cells below twenty movements are suppressed; twenty to forty-nine are descriptive and not ranked.
Source: CAA punctuality statistics · snapshot 17/08/2026 10:24 · method route-record-v2.1 · Versioned claims, formulas and entity mapping.
0 operations (0.00%) excluded because their published percentages could not be reconciled to whole numbers. Excluded operations enter neither the numerator nor the denominator. Outcome figures and comparisons are unavailable when the excluded share exceeds fifteen percent.
The four numbers at a glance
13.2% of 1,056 movements were disrupted over July 2025 to June 2026.
Average delay covers delayed movements only. Its range reflects rounded source means and the unknown count of exactly punctual movements.
| Disruption | Short-notice cancellations | Delay among delayed movements | Over three hours |
|---|---|---|---|
| 13.2% of 1,056 movements | no events in 1,056 movements | 25.2–61.6 minutes; delayed denominator 139–299 | 0.3% of 1,056 movements |
Definitions, denominators and limits
Arrivals into UK reporting airports only unless labelled departures. Denominator: operated movements plus short-notice cancellations (M+C). A movement is not a scheduled flight. Unmatched records are data quality, not outcomes. Cells below twenty movements are suppressed; twenty to forty-nine are descriptive and not ranked.
When late, how late?
0.3% of 1,056 movements passed the three-hour mark.
The ladder shows how many movements crossed each threshold. Cancellations are separate: no events in 1,056 movements.
| Delay | Share and sample | One event per movements |
|---|---|---|
| More than 15 minutes | 13.2% of 1,056 movements | 7.6 |
| More than 30 minutes | 7.3% of 1,056 movements | 13.7 |
| More than 60 minutes | 2.5% of 1,056 movements | 40.6 |
| More than 120 minutes | 0.9% of 1,056 movements | 105.6 |
| More than 180 minutes | 0.3% of 1,056 movements | 352.0 |
| More than 360 minutes | 0.1% of 1,056 movements | 1,056.0 |
Definitions, denominators and limits
Arrivals into UK reporting airports only unless labelled departures. Denominator: operated movements plus short-notice cancellations (M+C). A movement is not a scheduled flight. Unmatched records are data quality, not outcomes. Cells below twenty movements are suppressed; twenty to forty-nine are descriptive and not ranked.
Slightly late or occasionally a disaster?
0.9% of 1,056 movements passed two hours.
The histogram shows the published delay bands. Bucket shares use all reported operations, including unmatched records; cancellations are reported separately.
Definitions, denominators and limits
Arrivals into UK reporting airports only unless labelled departures. Denominator: operated movements plus short-notice cancellations (M+C). A movement is not a scheduled flight. Unmatched records are data quality, not outcomes. Cells below twenty movements are suppressed; twenty to forty-nine are descriptive and not ranked.
Better or worse since the start of the archive?
Best month: Feb 2025, 2.9% of 70 movements. Worst month: Oct 2023, 68.2% of 66 movements.
The line traces every month. Lower disruption is better; gaps are never zeros. Pandemic readings describe a different operating regime.
Every monthly observation
| Month | Disruption on arrivals |
|---|---|
| Jan 2018 | 17.1% of 41 movements — small sample, not ranked |
| Feb 2018 | no events in 34 movements — small sample, not ranked |
| Mar 2018 | 12.8% of 39 movements — small sample, not ranked |
| Apr 2018 | 46.7% of 30 movements — small sample, not ranked |
| May 2018 | 12.5% of 32 movements — small sample, not ranked |
| Jun 2018 | 30.2% of 43 movements — small sample, not ranked |
| Jul 2018 | 28.0% of 50 movements |
| Aug 2018 | 42.6% of 54 movements |
| Sep 2018 | 49.0% of 51 movements |
| Oct 2018 | 40.0% of 50 movements |
| Nov 2018 | 18.8% of 48 movements — small sample, not ranked |
| Dec 2018 | 17.6% of 51 movements |
| Jan 2019 | 8.2% of 49 movements — small sample, not ranked |
| Feb 2019 | 22.5% of 40 movements — small sample, not ranked |
| Mar 2019 | 25.5% of 51 movements |
| Apr 2019 | 50.0% of 50 movements |
| May 2019 | 51.0% of 49 movements — small sample, not ranked |
| Jun 2019 | 26.0% of 50 movements |
| Jul 2019 | 17.7% of 62 movements |
| Aug 2019 | 27.1% of 59 movements |
| Sep 2019 | 25.4% of 59 movements |
| Oct 2019 | 58.5% of 53 movements |
| Nov 2019 | 41.2% of 51 movements |
| Dec 2019 | 30.5% of 59 movements |
| Jan 2020 | 33.3% of 60 movements |
| Feb 2020 | suppressed: fewer than twenty movements |
| Mar 2020 | suppressed: fewer than twenty movements |
| Apr 2020 | suppressed: fewer than twenty movements |
| May 2020 | suppressed: fewer than twenty movements |
| Jun 2020 | suppressed: fewer than twenty movements |
| Jul 2020 | suppressed: fewer than twenty movements |
| Aug 2020 | suppressed: fewer than twenty movements |
| Sep 2020 | suppressed: fewer than twenty movements |
| Oct 2020 | suppressed: fewer than twenty movements |
| Nov 2020 | suppressed: fewer than twenty movements |
| Dec 2020 | suppressed: fewer than twenty movements |
| Jan 2021 | suppressed: fewer than twenty movements |
| Feb 2021 | suppressed: fewer than twenty movements |
| Mar 2021 | suppressed: fewer than twenty movements |
| Apr 2021 | suppressed: fewer than twenty movements |
| May 2021 | suppressed: fewer than twenty movements |
| Jun 2021 | suppressed: fewer than twenty movements |
| Jul 2021 | suppressed: fewer than twenty movements |
| Aug 2021 | suppressed: fewer than twenty movements |
| Sep 2021 | suppressed: fewer than twenty movements |
| Oct 2021 | suppressed: fewer than twenty movements |
| Nov 2021 | suppressed: fewer than twenty movements |
| Dec 2021 | suppressed: fewer than twenty movements |
| Jan 2022 | no data |
| Feb 2022 | no data |
| Mar 2022 | suppressed: fewer than twenty movements |
| Apr 2022 | no data |
| May 2022 | no data |
| Jun 2022 | no data |
| Jul 2022 | no data |
| Aug 2022 | suppressed: fewer than twenty movements |
| Sep 2022 | suppressed: fewer than twenty movements |
| Oct 2022 | suppressed: fewer than twenty movements |
| Nov 2022 | suppressed: fewer than twenty movements |
| Dec 2022 | suppressed: fewer than twenty movements |
| Jan 2023 | suppressed: fewer than twenty movements |
| Feb 2023 | suppressed: fewer than twenty movements |
| Mar 2023 | suppressed: fewer than twenty movements |
| Apr 2023 | 72.0% of 25 movements — small sample, not ranked |
| May 2023 | 52.9% of 34 movements — small sample, not ranked |
| Jun 2023 | 50.0% of 54 movements |
| Jul 2023 | 35.5% of 62 movements |
| Aug 2023 | 42.9% of 63 movements |
| Sep 2023 | 38.5% of 65 movements |
| Oct 2023 | 68.2% of 66 movements |
| Nov 2023 | 28.1% of 57 movements |
| Dec 2023 | 34.7% of 75 movements |
| Jan 2024 | 20.0% of 80 movements |
| Feb 2024 | 40.5% of 74 movements |
| Mar 2024 | 26.7% of 75 movements |
| Apr 2024 | 57.7% of 78 movements |
| May 2024 | 22.1% of 77 movements |
| Jun 2024 | 27.7% of 83 movements |
| Jul 2024 | 25.8% of 93 movements |
| Aug 2024 | 31.2% of 93 movements |
| Sep 2024 | 21.3% of 89 movements |
| Oct 2024 | 38.3% of 81 movements |
| Nov 2024 | 18.2% of 66 movements |
| Dec 2024 | 9.5% of 74 movements |
| Jan 2025 | 20.0% of 85 movements |
| Feb 2025 | 2.9% of 70 movements |
| Mar 2025 | 10.0% of 70 movements |
| Apr 2025 | 16.7% of 78 movements |
| May 2025 | 23.3% of 73 movements |
| Jun 2025 | 13.7% of 73 movements |
| Jul 2025 | 9.7% of 93 movements |
| Aug 2025 | 11.8% of 93 movements |
| Sep 2025 | 12.2% of 90 movements |
| Oct 2025 | 3.3% of 90 movements |
| Nov 2025 | 20.5% of 78 movements |
| Dec 2025 | 17.6% of 91 movements |
| Jan 2026 | 11.5% of 87 movements |
| Feb 2026 | 9.2% of 76 movements |
| Mar 2026 | 10.6% of 85 movements |
| Apr 2026 | 23.3% of 90 movements |
| May 2026 | 17.2% of 93 movements |
| Jun 2026 | 11.1% of 90 movements |
Definitions, denominators and limits
Arrivals into UK reporting airports only unless labelled departures. Denominator: operated movements plus short-notice cancellations (M+C). A movement is not a scheduled flight. Unmatched records are data quality, not outcomes. Cells below twenty movements are suppressed; twenty to forty-nine are descriptive and not ranked.
What did the big disruptions do here?
Each event is compared with the twelve preceding calendar months and the first later return to baseline.
A worse month is not proof the named event caused it. The open data carry no delay-cause field. Recovery is observed for at most eight subsequent months.
| Event | Direction | Disruption that month | Baseline before | Return to baseline |
|---|---|---|---|---|
| Pandemic regime · Jun 2020 | Arrivals | suppressed: fewer than twenty movements | no complete baseline | not enough data: needs a complete baseline and a publishable event month |
| Pandemic regime · Jun 2020 | Departures | suppressed: fewer than twenty movements | no complete baseline | not enough data: needs a complete baseline and a publishable event month |
| Summer 2022 breakdown · Jul 2022 | Arrivals | no data | no complete baseline | not enough data: needs a complete baseline and a publishable event month |
| Summer 2022 breakdown · Jul 2022 | Departures | no data | no complete baseline | not enough data: needs a complete baseline and a publishable event month |
| NATS system failure · Aug 2023 | Arrivals | 42.9% of 63 movements | no complete baseline | not enough data: needs a complete baseline and a publishable event month |
| NATS system failure · Aug 2023 | Departures | 9.5% of 63 movements | no complete baseline | not enough data: needs a complete baseline and a publishable event month |
| Heathrow fire closure · Mar 2025 | Arrivals | 10.0% of 70 movements | 25.1% across 964 flights | returned the next month |
| Heathrow fire closure · Mar 2025 | Departures | 4.3% of 70 movements | 12.1% across 962 flights | returned the next month |
Definitions, denominators and limits
Arrivals into UK reporting airports only unless labelled departures. Denominator: operated movements plus short-notice cancellations (M+C). A movement is not a scheduled flight. Unmatched records are data quality, not outcomes. Cells below twenty movements are suppressed; twenty to forty-nine are descriptive and not ranked.
Which time of year has been hardest?
Highest median: Apr, 48.3% across 6 months. Lowest median: Feb, 9.2% across 5 months.
Even the same calendar month differs across years. This is a record, not a forecast; pandemic years are excluded from the summaries.
Each cell is one calendar month of one year; the number is the share of that month’s arrivals that were disrupted. Darker means worse. A blank cell means no published data, not zero.
Month-by-month table: median, observed range and sample
| Month | Median | Observed range | Sample |
|---|---|---|---|
| Jan | 17.1% | 8.2–20.0% | 5 months, 342 flights |
| Feb | 9.2% | 0.0–40.5% | 5 months, 294 flights |
| Mar | 12.8% | 10.0–26.7% | 5 months, 320 flights |
| Apr | 48.3% | 16.7–72.0% | 6 months, 351 flights |
| May | 22.7% | 12.5–52.9% | 6 months, 358 flights |
| Jun | 26.9% | 11.1–50.0% | 6 months, 393 flights |
| Jul | 25.8% | 9.7–35.5% | 5 months, 360 flights |
| Aug | 31.2% | 11.8–42.9% | 5 months, 362 flights |
| Sep | 25.4% | 12.2–49.0% | 5 months, 354 flights |
| Oct | 40.0% | 3.3–68.2% | 5 months, 340 flights |
| Nov | 20.5% | 18.2–41.2% | 5 months, 300 flights |
| Dec | 17.6% | 9.5–34.7% | 5 months, 350 flights |
Definitions, denominators and limits
Arrivals into UK reporting airports only unless labelled departures. Denominator: operated movements plus short-notice cancellations (M+C). A movement is not a scheduled flight. Unmatched records are data quality, not outcomes. Cells below twenty movements are suppressed; twenty to forty-nine are descriptive and not ranked.
Which routes have the best and worst record?
Lowest disruption: LGW ↔ ZHENGZHOU-XINZHENG, 3.8% of 105 movements. Highest disruption: LHR ↔ GUANGZHOU-BAIYUN-INTERNATIONAL, 20.0% of 375 movements.
The airline table uses this carrier on each route; airport and country tables combine carriers. Each figure is arrivals into its UK airport. Only sufficiently large exact samples enter the ordered comparison.
All route comparisons and sample flags
| Route | Disruption | Cancellations | Sample flag |
|---|---|---|---|
| LGW ↔ ZHENGZHOU-XINZHENG | 3.8% of 105 movements | no events in 105 movements | descriptive comparison |
| LHR ↔ WUHAN-TIANHE-INTERNATIONAL | 3.8% of 104 movements | no events in 104 movements | descriptive comparison |
| LHR ↔ BEIJING-DAXING-INTERNATIONAL-AIRPORT | 9.6% of 365 movements | no events in 365 movements | descriptive comparison |
| LGW ↔ GUANGZHOU-BAIYUN-INTERNATIONAL | 19.6% of 107 movements | no events in 107 movements | descriptive comparison |
| LHR ↔ GUANGZHOU-BAIYUN-INTERNATIONAL | 20.0% of 375 movements | no events in 375 movements | descriptive comparison |
Definitions, denominators and limits
Arrivals into UK reporting airports only unless labelled departures. Denominator: operated movements plus short-notice cancellations (M+C). A movement is not a scheduled flight. Unmatched records are data quality, not outcomes. Cells below twenty movements are suppressed; twenty to forty-nine are descriptive and not ranked.
Which UK airports appear in this record?
2 published groups have enough arrivals to describe.
These operations are not matched for season, aircraft or time of day. The table makes no causal comparison.
| UK airport | Disruption | Sample flag |
|---|---|---|
| Gatwick | 11.8% of 212 movements | descriptive comparison |
| Heathrow | 13.5% of 844 movements | descriptive comparison |
Definitions, denominators and limits
Arrivals into UK reporting airports only unless labelled departures. Denominator: operated movements plus short-notice cancellations (M+C). A movement is not a scheduled flight. Unmatched records are data quality, not outcomes. Cells below twenty movements are suppressed; twenty to forty-nine are descriptive and not ranked.
Arrivals vs departures
13.2% of 1,056 movements arrivals were disrupted, against 9.0% of 1,060 movements departures.
The directions have separate denominators. A departure delay does not establish late arrival at the final destination; legal conditions differ by direction and carrier.
Definitions, denominators and limits
Arrivals into UK reporting airports only unless labelled departures. Denominator: operated movements plus short-notice cancellations (M+C). A movement is not a scheduled flight. Unmatched records are data quality, not outcomes. Cells below twenty movements are suppressed; twenty to forty-nine are descriptive and not ranked.
Movements past the three-hour mark
0.3% of 1,056 movements were more than three hours late over July 2025 to June 2026.
Cancellation is counted separately and is not a three-hour delay. Monthly counts cannot establish a passenger claim.
Definitions, denominators and limits
Arrivals into UK reporting airports only unless labelled departures. Denominator: operated movements plus short-notice cancellations (M+C). A movement is not a scheduled flight. Unmatched records are data quality, not outcomes. Cells below twenty movements are suppressed; twenty to forty-nine are descriptive and not ranked.
Facts about this record
China Southern appears from Jan 2018 to Jun 2026 with 4,295 arrivals.
The bounds describe the published record, not every commercial service. Final point is not necessarily a nonstop destination.
Archive route pairs: 13.
Definitions, denominators and limits
Arrivals into UK reporting airports only unless labelled departures. Denominator: operated movements plus short-notice cancellations (M+C). A movement is not a scheduled flight. Unmatched records are data quality, not outcomes. Cells below twenty movements are suppressed; twenty to forty-nine are descriptive and not ranked.
What this data cannot tell you
Delay cause, individual passenger experience, commercial decisions and future performance are not established here.
Also absent are aircraft type, within-month time-of-day splits and cancellations made ahead of the short-notice reporting window.
0 operations (0.00%) excluded because their published percentages could not be reconciled to whole numbers. Excluded operations enter neither the numerator nor the denominator. Outcome figures and comparisons are unavailable when the excluded share exceeds fifteen percent.
Definitions, denominators and limits
Arrivals into UK reporting airports only unless labelled departures. Denominator: operated movements plus short-notice cancellations (M+C). A movement is not a scheduled flight. Unmatched records are data quality, not outcomes. Cells below twenty movements are suppressed; twenty to forty-nine are descriptive and not ranked.
Quick answers
These answers use the same populations and window as the tables above.
Every percentage retains its denominator.
How disrupted are these arrivals?
13.2% of 1,056 movements over July 2025 to June 2026.
How often are they cancelled?
no events in 1,056 movements at short notice.
How late is late?
25.2–61.6 minutes; delayed denominator 139–299
Are services improving?
The monthly line records past changes; it is not a forecast or a causal comparison.
Which routes are compared?
The route table uses separately identified UK airport and final-point pairs, with small samples flagged.
Does this cover my flight?
Only published aggregate movements; the circumstances of any individual flight are unknown.
Definitions, denominators and limits
Arrivals into UK reporting airports only unless labelled departures. Denominator: operated movements plus short-notice cancellations (M+C). A movement is not a scheduled flight. Unmatched records are data quality, not outcomes. Cells below twenty movements are suppressed; twenty to forty-nine are descriptive and not ranked.
How this was calculated
Source: CAA punctuality statistics · snapshot 17/08/2026 10:24 · method route-record-v2.1 · Versioned claims, formulas and entity mapping.
Integer bucket counts are recovered from rounding intervals over D=M+C+U with their sum constrained to M. Ambiguous counts remain ranges; failed reconciliation is quarantined.
Definitions, denominators and limits
Arrivals into UK reporting airports only unless labelled departures. Denominator: operated movements plus short-notice cancellations (M+C). A movement is not a scheduled flight. Unmatched records are data quality, not outcomes. Cells below twenty movements are suppressed; twenty to forty-nine are descriptive and not ranked.
Your trip has already gone wrong
The monthly line and events table show whether your month was typical or unusual in the published record.
Nothing here decides an individual case or what a passenger may be owed.
CAA: the official route to complainDefinitions, denominators and limits
Arrivals into UK reporting airports only unless labelled departures. Denominator: operated movements plus short-notice cancellations (M+C). A movement is not a scheduled flight. Unmatched records are data quality, not outcomes. Cells below twenty movements are suppressed; twenty to forty-nine are descriptive and not ranked.