中文
中文版

Taiwan Earthquake and Weather Relationship Study

A Decade of Data-Driven Analysis (2016-2026) | 206 Earthquakes M5.0+ | 2026-04-05
USGS Earthquake Database Open-Meteo Weather API Claude Opus 4.6 Analysis

Taiwan experienced 206 earthquakes of magnitude 5.0 or higher in a decade.
We compared the weather data for the 15 days before each major earthquake to see whether any relationship exists between weather and earthquakes.
Table of Contents
  1. Methodology
  2. Earthquake Overview: 206 Events M5.0+ in Ten Years
  3. M6.0+ Major Earthquake List
  4. Weather Data for 15 Days Before Earthquakes
  5. Experimental Group vs. Control Group Comparison
  6. Five Key Observations
  7. Noteworthy Extreme Cases
  8. Conclusions and Future Research Directions

1. Methodology

Data Sources

  • Earthquake data: USGS Earthquake Catalog API
  • Weather data: Open-Meteo Historical Weather API
  • Period: 2016/01/01 — 2026/04/05
  • Geographic scope: Lat 21.5°-25.5°N, Lon 119.5°-122.5°E

Analytical Method

  • Experimental group: 18 M6.0+ earthquakes, weather for 15 days prior
  • Control group: 18 randomly selected "no major earthquake" dates, same 15-day weather window
  • Weather stations: Nearest station to earthquake location (Hualien/Yilan/Tainan/Taitung)
  • Comparison metrics: Pressure, temperature, rainfall, rainy days, humidity, pressure range

Disclaimer: This is an exploratory data analysis, not a rigorous academic paper. Sample size is limited (only 18 M6.0+ events), and conclusions are for reference only and do not constitute prediction criteria.

2. Earthquake Overview: 206 Events M5.0+ in Ten Years

206
Total M5.0+ Earthquakes
19
M6.0+ Major Earthquakes
7.4
Maximum Magnitude
(2024/04/02 Hualien)

Annual Distribution

2016
9
2017
4
2018
15
2019
14
2020
8
2021
23
2022
42
2023
4
2024
67
2025
18
2026
2*

*2026 data through April 5. The 2024 spike is primarily from the April 3 Hualien earthquake and its numerous aftershocks.

Monthly Distribution (All 206 Events)

Jan
15
Feb
22
Mar
14
Apr
71
May
14
Jun
15
Jul
6
Aug
8
Sep
16
Oct
7
Nov
8
Dec
10

The April spike (71 events) is primarily from the 2024 April 3 Hualien earthquake sequence. Excluding April 2024, the monthly distribution is more even, though winter-spring (Dec-Apr) still shows slightly higher counts.

3. M6.0+ Major Earthquake List

DateMagDepthLocationNotes
2024-04-027.440 kmHualienDecade's largest, April 3 Hualien earthquake
2022-09-186.910 kmTaitungSep 18 Chishang earthquake
2022-03-226.724 kmHualien
2025-12-276.663 kmYilan
2022-09-176.510 kmTaitungSep 18 foreshock
2016-02-056.423 kmTainanFeb 6 Meinong earthquake
2018-02-066.417 kmHualienFeb 6 Hualien earthquake
2024-04-036.414 kmHualienApril 3 aftershock
2021-10-246.269 kmYilanDeep earthquake
2022-01-036.219 kmHualien
2022-05-096.221 kmHualien offshore
2018-02-046.112 kmHualienFeb 6 foreshock
2019-04-186.120 kmHualien
2020-12-106.171 kmYilanDeep earthquake
2024-04-226.110 kmHualienApril 3 aftershock sequence
2024-04-226.19 kmHualienSame-day double quake
2024-08-156.114 kmHualien
2025-01-206.016 kmTainan
2025-12-246.08 kmTaitung

4. Weather Data for 15 Days Before M6.0+ Earthquakes

Below is the weather summary for the 15 days (day -15 to day -1) preceding each M6.0+ earthquake:

Quake DateMagLocationAvg Temp
°C
Total Rain
mm
Rainy
Days
Heavy
Rain
Avg Press
hPa
Press Range
hPa
Avg Humid
%
2016-02-056.4Tainan15.1125.41021022.212.291.1
2018-02-046.1Hualien18.6120.11501018.910.284.7
2018-02-066.4Hualien17.5138.71501020.010.783.5
2019-04-186.1Hualien24.674.11301014.17.085.3
2020-12-106.1Yilan20.6254.91551022.05.885.2
2021-10-246.2Yilan25.8319.81251014.415.684.9
2022-01-036.2Hualien20.129.3501022.313.780.7
2022-03-226.7Hualien23.133.0501014.510.780.3
2022-05-096.2Hualien26.875.0811012.78.484.9
2022-09-176.5Taitung22.9155.91321009.017.891.5
2022-09-186.9Taitung22.8142.91221009.517.891.3
2024-04-027.4Hualien23.316.7701017.013.682.2
2024-04-036.4Hualien23.814.2601016.613.681.4
2024-04-226.1Hualien26.927.1601013.08.682.3
2024-08-156.1Hualien31.128.6701007.15.979.5
2025-01-206.0Tainan15.913.1501020.79.579.3
2025-12-246.0Taitung19.311.3301019.13.079.8
2025-12-276.6Yilan20.968.6901020.39.281.9

5. Experimental Group vs. Control Group Comparison

The control group consists of 18 randomly selected "no major earthquake" dates (evenly distributed across 2016-2025 seasons), using the same 15-day weather window.

Avg Pressure
vs
1016.3 hPa ← Pre-quake | Control → 1013.5 hPa
Pressure Range
vs
10.7 hPa ← Pre-quake | Control → 9.3 hPa
15-Day Rainfall
vs
91.6 mm ← Pre-quake | Control → 59.3 mm (+54%)
Rainy Days
vs
9.2 days ← Pre-quake | Control → 7.4 days (+24%)
Avg Humidity
vs
83.9% ← Pre-quake | Control → 80.1% (+3.8%)
Avg Temperature
vs
22.2°C ← Pre-quake | Control → 26.5°C (-4.3°C)
High Humidity Rate
(>80%)
vs
83% (15/18) ← Pre-quake | Control → 56% (10/18)

6. Five Key Observations

Finding 1: Major Earthquakes Prefer Winter-Spring

Of the 18 M6.0+ earthquakes, 8 occurred in winter (44%) and 6 in spring (33%), totaling 78%. Only 1 occurred in summer (6%).

This implies that differences in temperature, pressure, and rainfall may simply reflect seasonal effects rather than independent weather factors.

However, the seismological community remains divided on whether earthquakes exhibit seasonality. Some studies suggest the crust is influenced by tidal forces and atmospheric pressure loading, and winter high-pressure systems may impose additional stress on shallow faults.

Finding 2: Pre-Earthquake Rainfall Is Elevated (+54%)

Average rainfall in the 15 days before earthquakes was 91.6mm, 54% higher than the control group's 59.3mm. This is the most prominent difference.

Even accounting for seasonal factors, some cases had exceptionally high rainfall: 254.9mm before the Dec 2020 Yilan M6.1 and 319.8mm before the Oct 2021 Yilan M6.2.

Academic context: "Rainfall-induced seismicity" is a known geophysical phenomenon. Heavy precipitation infiltrating the ground increases pore water pressure, potentially reducing the effective normal stress on faults and making faults already near critical state more likely to slip. This effect is more pronounced for shallow earthquakes (<20km).

Finding 3: Pre-Earthquake Humidity Is Elevated (83.9% vs. 80.1%)

83% (15/18) of major earthquakes had average humidity above 80% in the preceding 15 days, compared to only 56% in the control group.

However, Taiwan is inherently a high-humidity environment, and the northeast monsoon brings moisture in winter-spring, so a larger sample is needed to confirm significance.

Finding 4: Pressure Range May Be the More Interesting Indicator

The 15-day pressure range (max-min) averaged 10.7 hPa before earthquakes vs. 9.3 hPa for controls. The gap isn't large, but several major earthquakes showed notably high pressure ranges:

Academic context: The hypothesis that atmospheric pressure changes can trigger earthquakes has a long history. Rapid pressure changes (such as passing typhoon low-pressure systems) can theoretically exert small but measurable forces on the crust. The Sep 2022 Taitung earthquake was preceded by a typhoon approaching Taiwan.

Finding 5: 0% of Major Earthquakes Occurred During "Drought Periods"

Among the 18 M6.0+ earthquakes, not a single one had less than 10mm of total rainfall in the preceding 15 days. The control group also had none (Taiwan rarely goes 15 days without rain), but the driest pre-earthquake periods still recorded 11-16mm.

From another angle: Taiwan's major earthquakes never occur during completely dry periods. This "absence" itself is a pattern worth noting.

7. Noteworthy Extreme Cases

Case 1: Oct 24, 2021 Yilan M6.2 — Major Earthquake After Heavy Rain

15-day pre-quake rainfall of 319.8mm (highest of 18 events), including 5 days of heavy rain (>20mm/day). Pressure range 15.6 hPa. This is the case most consistent with the "rainfall-induced earthquake" hypothesis.

However, this earthquake had a depth of 69km (deep), and rainfall infiltration theoretically cannot affect such depths.

Case 2: Sep 17-18, 2022 Taitung M6.5+M6.9 — Typhoon Season Double Quake

15-day rainfall 155.9mm, pressure range 17.8 hPa (highest), humidity 91.5% (highest). During this period, a typhoon system was approaching Taiwan, causing dramatic pressure and weather changes. Two M6+ earthquakes occurred within two days.

Case 3: Apr 2, 2024 Hualien M7.4 — Decade's Largest, Unremarkable Weather

15-day pre-quake rainfall was only 16.7mm, with relatively calm weather. Pressure range 13.6 hPa was slightly elevated but not extreme. This is the best counterexample for weather-earthquake independence — the decade's largest earthquake occurred during the calmest weather.

8. Conclusions and Future Research Directions

Conclusion: No Causal Weather-Earthquake Relationship Can Be Established

The observed differences (54% more rainfall, 3.8% higher humidity, 4.3°C lower temperature) are most likely primarily due to seasonal bias — Taiwan's major earthquakes statistically prefer winter-spring, which naturally features low temperatures, high humidity, and more rainfall.

The strongest counterevidence is the 2024 Hualien M7.4 (decade's strongest) occurring during extremely calm weather, demonstrating that weather is not a necessary condition for major earthquakes.

Two Directions Worth Further Investigation

  1. Extreme rainfall and shallow earthquake correlation
    "Rainfall-induced seismicity" is a known mechanism in academia (pore water pressure effect). The Dec 2020 and Oct 2021 cases in this study do show abnormally high rainfall + earthquake combinations. Recommendation: isolate shallow earthquakes (depth < 20km) and independently analyze their correlation with preceding rainfall.
  2. Pressure change time series and earthquakes
    Several major earthquakes were preceded by high pressure ranges (especially during typhoon season). Recommendation: instead of 15-day averages, construct daily pressure change curves to observe whether abnormal pressure fluctuations occur 1-3 days before earthquakes.

Study Limitations

Data sources: USGS Earthquake Catalog API | Open-Meteo Historical Weather API | Central Weather Administration Seismological Center