IBTrACS v04r01 · NI Basin · 1842–2025

Into the Eye of the Bay

A data-driven exploration of 183 years of tropical cyclones over the Bay of Bengal — where storms are born, how they move, and what the long record reveals about a changing ocean.

10 VISUALISATIONS · STORM TRACKS · GENESIS · INTENSITY · LANDFALL · CLIMATOLOGY

scroll
183
Years of Record
~600+
Named Storms
44 m/s
Peak Wind on Record
Nov
Peak Activity Month
01

Track Density

Every Storm, Every Track —
183 Years Layered

The kernel density overlay of all IBTrACS North Indian Basin tracks reveals the Bay of Bengal's circulation anatomy. The hottest corridor arcs north-northeast from ~10°N toward the Bangladesh-Myanmar coast, where storm convergence is densest. Yellow-orange threads mark the most intense systems slicing through the central Bay.

Bay of Bengal Cyclone Track Density Map 1842-2025
Convergence hotspot sits around 20–22°N near the head of the Bay — almost every track eventually recurves or terminates here toward the Bangladesh delta or Odisha coast.
Eastern corridor divergence is evident off the Andaman Islands (∼90–95°E), where some storms are spawned and propagate westward at lower latitudes before intensifying.
Tropical Depression (TD) tracks in cyan form a diffuse background; Cat 1+ systems in gold-orange follow tighter, more coherent paths — a direct signature of steering-flow organisation at higher intensities.
02

Annual Trend 1970–2025

More Storms, but Moderate Intensity

The 55-year record tells a clear story: storm frequency has roughly tripled since the 1970s, yet average peak intensity has not kept pace. The green dashed linear trend in storm counts climbs steadily from ~0.5/yr to nearly 4/yr — while the orange intensity curve oscillates without a strong upward signal after 2000.

Annual Storm Frequency and Intensity Trend 1970-2025
Frequency surge post-2005: The 5-yr rolling mean jumps from ~1 storm/yr in the 1990s to 3–4/yr after 2005, coinciding with improved satellite detection and genuinely elevated basin activity.
Intensity plateau: Average peak winds hovered 20–25 kt through most of the record with no sustained upward trend — suggesting frequency, not intensity, is the primary change signal in the NI Basin.
Active decades: 2010s stand out with the highest annual counts (up to 6), while early 2020s maintain elevated frequency but show a slight intensity dip.
03

Seasonal Cycle

A Tale of Two Peaks:
May and November

Unlike the Atlantic basin with a single September peak, the Bay of Bengal has a bimodal season. Post-monsoon (October–December) strongly dominates, with November as the absolute maximum. A secondary pre-monsoon peak emerges in April–May, bookending the June–September monsoon suppression period when wind shear inhibits cyclogenesis.

Bay of Bengal Seasonal Distribution and Cyclone Rose
November reigns with the highest observation count (~4,800) — nearly twice October and three times the monsoon months. The cyclone rose (right panel) confirms the dominant Nov–Oct–Dec arc visually.
Monsoon suppression in June–August is real but not absolute: the rose shows meaningful activity in Jul–Sep (yellow sector), mostly weaker systems that form in favourable monsoon troughs.
January–March are near-dormant, with observation counts below 400 per month — a risk window that is easy to overlook in preparedness planning.
04

Intensity Distribution

Most Observations Are Weak —
the Tail is What Kills

The wind-speed distribution is sharply right-skewed: the mode sits below 10 m/s (tropical depression strength), and the curve drops steeply toward Cat 1+ thresholds. Yet the right tail — rare but deadly — represents all major landfalling disasters. The decade stacked bars confirm that TD-class (blue) dominates every era.

Wind Speed Distribution and Intensity Category by Decade
TD class overwhelms — the teal bar in each decade's stacked chart is barely visible compared to the blue TD fraction, which accounts for ~85% of all observations across all periods.
Decade-to-decade totals vary: the 1970s and 1990s show relatively high observation counts, with the 2020s lower simply because the decade is incomplete (data through 2025).
Cat 3+ observations (red-purple) are vanishingly rare in the bar chart, underscoring the Bay's reputation for producing frequent but rarely extreme storms compared to the western Pacific.
05

Genesis Points

Where Storms Are Born

Genesis points — the first recorded position of each storm — tell us where atmospheric and oceanic conditions conspire to spin up circulation. The Bay shows a striking clustering of Oct–Dec genesis (orange) across the southern and central basin, while Apr–Jun storms (green) emerge slightly further south. Larger markers signal higher peak speeds.

Cyclone Genesis Points Bay of Bengal

The dense cluster along the eastern coast (~90–95°E, 10–15°N) represents storms generated near the Andaman Sea and northern Indian Ocean — a known nursery tied to the Inter-Tropical Convergence Zone and monsoonal moisture flux.

Post-monsoon genesis (orange, Oct–Dec) dominates in both count and geographic spread, extending from 5°N all the way to 20°N — confirming the broad latitudinal reach of winter cyclogenesis.
Pre-monsoon storms (green, Apr–Jun) preferentially form in the southern Bay between 5–15°N, where warm sea-surface temperatures build before the summer monsoon onset.
Larger, stronger storms (big markers) appear clustered in the mid-basin and eastern coast, suggesting that storms forming with room to develop over open water reach higher peak intensities.
06

Landfall Analysis

August is the Deadliest
Landfall Month

While November produces the most storms at sea, August registers the peak landfall count (281 episodes). This counterintuitive split reflects that monsoon-season storms, though weaker, tend to track directly into coastlines — while post-monsoon systems often recurve or dissipate over the open ocean.

Landfall Distance to Coastline and Landfall Episodes by Month
Distance-to-land distribution (left panel) shows a spike near 500 km — the typical open-ocean position — and a secondary peak near the coast (0–20 km), reflecting the high frequency of coastal observations from landfalling systems.
June–September dominates landfall counts (blue bars) with Aug–Jul–Jun in the top three, closely followed by the post-monsoon Oct–Nov season (orange bars) which brings more intense storms ashore.
The 200 km buffer line helps contextualise the distance distribution — most offshore observations sit well beyond this threshold, meaning "close-call" near-coastal storms are a distinct, trackable subset of the record.
Key Finding

Frequency of landfall and intensity of landfall are driven by different seasons. Bangladesh, Odisha, and Andhra Pradesh face their highest landfall probability in August, but their highest landfall intensity exposure in October–November.

07

Decade × Month Heatmap

How the Season Has Shifted
Over 120 Years

Each cell shows the count of unique named storms per month-decade combination. The pre-satellite era (pre-1960s) is sparse and uniform, but post-2000 cells light up dramatically — especially May, October, and November in the 2010s, which glow yellow-orange with 6–8 storms per cell.

Storm Frequency Heatmap Decade by Month
2010s stands out starkly: October (7), November (8), and December (6) all reach their historical peak — signaling a possible intensification of the post-monsoon season in recent decades.
May activity is a recent phenomenon: pre-1990, May cells are almost universally "1". By the 2000s and 2010s, May jumps to 4 and 6 storms — possibly linked to earlier ocean warming in the spring.
Data sparsity in early decades (1900s–1940s) must be interpreted cautiously — the near-uniform "1" pattern reflects observation limitations, not genuinely low activity.
08

Record-Setters

The Top 20 Most
Intense Storms

An unnamed 1956 storm towers above all others at 44 m/s peak wind — classified as Category 1. Sitrang (2022) follows at 32 m/s, and Mahasen/Viyaru (2013) at 27 m/s. Remarkably, most "top 20" storms are Tropical Storm (TS) class, reflecting the Bay's character as a basin that spins up frequent but rarely super-intense systems.

The recent decade (2007–2023) contributes 11 of the top 20 entries, including Sidr (2007), Rashmi (2008), Aila (2009), and Midhili (2023) — suggesting improved observational records are capturing systems that would have been missed in earlier decades.

1956 unnamed storm is the only Cat 1 in the top 20 and remains an outlier — whether this reflects a genuine extreme event or a data artefact in the historical record requires scrutiny.
Sitrang (2022) is the most intense modern storm in the dataset at 32 m/s, making landfall in Bangladesh — a sobering reminder of contemporary risk.
Sidr (2007) at 22 m/s caused catastrophic flooding in Bangladesh despite its moderate peak wind — illustrating that surge and rainfall matter as much as wind speed for impact.
Top 20 Most Intense Storms NI Basin
09

Rapid Intensification

Sudden Surges in
Wind Speed

Rapid Intensification (RI) — defined here as a wind speed increase of ≥15 m/s per time step (~30 kt/24h) — is sporadic in the Bay of Bengal. The left panel shows RI events peaked around 1996–2000 and have trended slightly downward since, though data limitations in early decades complicate interpretation.

Rapid Intensification Events 1970-2025
RI events are rare but impactful: only a handful of years exceed 1–2 episodes, but each represents a storm that became significantly more dangerous than forecast within 24 hours.
The ΔSpeed vs Current Intensity scatter (right panel) shows that RI-threshold crossings (above the orange dashed line) occur most frequently in storms already at 10–20 m/s — moderate-intensity systems with the thermodynamic room to explode upward.
Recent years (2015+, yellow dots) appear more densely scattered at moderate intensities, potentially reflecting improved data resolution rather than a true increase in RI frequency.
Forecast Challenge

RI remains the hardest problem in tropical cyclone forecasting. Even 24-hour advance warning of an RI event can cut casualties dramatically — making these rare data points disproportionately valuable for model development.

10

Climatology Dashboard

The Full Picture —
Six Panels, One Basin

The summary dashboard synthesises the entire 55-year modern record into six complementary views. Together they paint a coherent portrait: a basin with surging storm counts, stable-to-declining average intensity, a dominant Tropical Storm classification, a landfall record skewed toward the 1970s–80s, and an October–December season that has grown markedly more active.

Bay of Bengal Cyclone Climatology Dashboard
Panel A (Storm Count) confirms the post-2005 frequency surge — the 5-yr gold average reaches 4+ storms/yr in the 2010s, nearly four times the 1970s baseline.
Panel B (Avg Peak Intensity) shows high variability with a notable spike around 2000 followed by a gradual decline — a pattern that may reflect changing storm populations rather than a true weakening of individual storms.
Panel C (TC-Class Fraction) reveals a data-classification shift in the early 1990s where the TS fraction drops sharply — a known artefact of changing agency reporting conventions in the IBTrACS dataset.
Panel E (Storm Nature): 78% of observations are Tropical Storm (TS) class; only 16% are classified NR (non-rotating disturbance), with MX and DS types marginal — confirming the Bay's strongly convective, warm-core storm character.
Panel F (Landfall by Decade): The 1970s saw the highest landfall count (~82 episodes), likely reflecting both active conditions and different classification thresholds. 2020s landfall is already at ~38 with five years remaining.

What the data tells us

The Bay of Bengal is Getting
Busier, Not Calmer

Storm frequency has risen sharply over the instrumental record. The post-monsoon season has grown markedly more active since the 2000s. While individual storm intensity has not uniformly increased, more storms means more chances for a destructive event. For over 400 million people living in the Bay's coastal zone — from Odisha to Myanmar — understanding these patterns is a matter of survival.

Data: IBTrACS v04r01 · NOAA/WMO · NI Basin 1842–2025