Early-stage metocean comparison across three sites along the Taiwan Strait
⏱ ~6 min read
This page summarises a methodology sample produced by PF Tech Services Ltd. Full PDF download at the bottom of this page.
"Over the 2019–2023 analysis window, the three sites show distinct characteristics across wind, wave, operability, and typhoon exposure — no single site is best across all four; a clear three-way trade-off."
Site location
Three sites: North (25.00°N, 120.44°E), Mid (24.05°N, 119.80°E), South (23.00°N, 119.57°E) — each approximately 50 km from the nearest coastal point of Taiwan.
▶ Executive Summary
At a glance
Three-way trade-off
The four dimensions show structural differences; no single site is best across all of them. North ranks highest in mean wind resource and (over 2019–2023) has the lowest typhoon passage frequency among the three sites. Mid — driven by the mid-strait channel funnelling effect (wind acceleration and directional concentration where the strait narrows) — has the highest wave extreme value (Hs p99, ~8% above North). South has the highest operability — at the standard tier, about 48 more standard operating days per year than the North.
5-year absolute extremes (Hs max, wind10 max) and typhoon frequency rankings are event-sample counts over 2019–2023, not climatologically stable values; they may not hold under resampling.
What we found
Wave extreme rankings differ from mean wind speed rankings
Annual mean wind speed decreases from north to south (North 8.91, Mid 8.42, South 6.77 m/s). Annual mean wave height Hs is nearly identical at North (1.60 m) and Mid (1.59 m), with South lower (1.35 m). But at the wave high end (Hs p99), the ranking reverses: Mid 4.66 m, North 4.33 m — Mid is about 8% higher than North. The cause is the mid-strait channel funnelling effect: the narrowing topography between two coasts accelerates and concentrates winds, occasionally producing larger waves.
South has the highest operability, but shows swell-dominated characteristics
South ranks highest across all three operability tiers. The median bad-weather event is shortest (15 h vs North 21 h, Mid 27 h), and the p90 event duration is also shortest (63 h vs North 90 h, Mid 106 h — i.e. 90% of South's bad-weather events end within 63 h). Annual mean wave height Hs is also lowest at South (1.35 m). "Bad weather" here means continuous periods when sea state exceeds operational thresholds and marine operations cannot proceed.
However, note that South's time-averaged wave energy is swell-dominated (swell accounts for 47.9% of time-averaged wave energy, exceeding wind-sea at 35.0%), and swell direction differs from local wind direction (swell direction ~138°, local wind from northeast). This describes the time-averaged energy structure, not the primary driver of extreme events or design loads — design loads from extreme events are still expected to be wind-sea driven. But period-sensitive operations (lifting, motion-sensitive vessels) at South will be more sensitive to wave period (Tp) variations than at the other two sites — a nuance not visible from Hs alone.
Typhoon exposure varies with latitude — within this 5-year sample
Over 2019–2023, 9 typhoons passed within 200 km of at least one site. Site frequencies: North 0.6/yr, Mid 1.2/yr, South 1.6/yr. North's lower frequency is consistent with Taiwan's terrain providing partial shielding at the strait's north exit; South has the highest exposure, with shorter median passage distance.
This is a statistical count over the 2019–2023 sample, not long-term climatology — 5 years is a sample snapshot, not a multi-decadal extreme-value estimate.
How we know
Mid-strait channel funnelling signature
Paired wind + wave roses at the three sites (2019–2023). Top row: ERA5 10 m wind direction (blue tones); bottom row: WAVERYS total wave direction (orange tones). All three sites show NNE-dominant winds; Mid has the highest directional concentration in the NNE sector — i.e. channel-funnelling-dominated winds, which in turn drive the wave extreme characteristic.
Typhoon tracks within 200 km of the three sites
The nine typhoons (IBTrACS catalogue) that passed within 200 km of any study site during 2019–2023. Track colour represents intensity at that point (Saffir-Simpson scale). Numbered labels are placed at each track's first entry point into the map, corresponding to the typhoon event list at the bottom. The three sites are marked with stars, with same-colour dashed circles outlining the 200 km filter range.
Operability persistence
Survival curves of operability persistence at the three sites (2019–2023). One panel per site, showing curves for three threshold tiers: sensitive, standard, robust. Log scale. The South curve extends further right at the top end (more long-duration operability windows); Mid has the longest single-event tail across all three tiers (a 2,535-hour outlier event in mid-2022, which does not change the median and p90 comparison).
Cross-validation against alternative-source 10-year wave hindcast
Monthly mean Hs cross-validation at the three sites: WAVERYS reanalysis (2019–2023, orange) vs Taiwan CWA WW3 climatology (2008–2017, blue), bilinearly interpolated to each site coordinate. Pearson correlation of the monthly seasonal cycle: North and Mid ≥ 0.97, South = 0.77, supporting the WAVERYS site climatology at the monthly seasonal-pattern level. This validation does not extend to absolute-magnitude verification — a ~0.25 m systematic bias exists between the three sites, which is non-negligible — nor does it constitute extreme-value validation.
ⓘ How to read this sample
This is a methodology sample, not a project deliverable. Intended for prospective customers evaluating PF Tech's "Early-stage Metocean Desktop Study (Level 1 Metocean Desktop Study)" service:
- What this report shows — the analytical depth, framing, and limitations discipline of PF Tech's early-stage desktop study, demonstrated across three sites along the Taiwan Strait.The methodology framework and report structure have long-term reference value; specific values are illustrative.
- What it is good for — multi-site early-stage screening and comparison, technology-selection scoping, and feasibility-stage operability discussions.
- What it is not — as a specific-project engineering input. The three selection points are methodology demonstrators, not actual project locations; values shown are not site-specific design figures.
The full report — methodology, complete data sources, full caveats, spectral partition analysis, CWA cross-validation, 15-item structured limitations chapter, and reproducibility manifest — PDF download below.
📄 See the full deliverable
Complete bilingual report — methodology, data sources, spectral partition analysis, CWA cross-validation, 15-item structured limitations, reproducibility manifest.
✉ Looking to commission a similar study?
Whether your project is offshore site assessment, port planning, or any other application requiring integrated metocean, wind, or typhoon-exposure analysis — we tailor delivery to your site needs (single or multi-site).
CONTACT
Or email us directly at contact@pft-s.com.tw
Disclaimer
This page summarises a methodology sample. It does not constitute a design basis for any specific wind farm, port facility, or coastal asset. PF Tech makes no representation about whether the three selection points fall within any specific operating or planned project area. The full Disclaimer is on page 3 of the PDF.
PF Tech Services Ltd. — 沛風科技服務有限公司. © 2026 PF Tech Services Ltd. All rights reserved.
Data attribution: ERA5 (CC-BY 4.0 Copernicus Climate Change Service), Copernicus Marine WAVERYS (CMEMS Licence), IBTrACS (NOAA NCEI public domain), Taiwan CWA OpenData (Open Government Data License v1). Full attribution and DOIs in §9 of the PDF.