23 Typhoons, 44 Million People Affected

Another typhoon warning. Another round of evacuations. Another news cycle of flooded streets, blown-off roofs, and donation drives. If you live in the Philippines long enough, this rhythm starts to feel normal — and that normalization is exactly what worries me.

I wanted to step back from the annual cycle and look at typhoon impact in aggregate. Not just the headline-grabbing ones like Yolanda, but all of them — the "minor" ones that still displace thousands, the ones that don't trend on social media but wreck provincial economies for months.

What I Built

An analysis of NDRRMC (National Disaster Risk Reduction and Management Council) data covering 23 typhoons between 2014 and 2020. The project tracks affected populations, infrastructure damage in pesos, agricultural losses, and regional impact. I normalized damage values for inflation and mapped which regions absorb the most punishment year after year.

Why This Project

Typhoon data in the Philippines is scattered across NDRRMC situation reports, news articles, and government press releases. There's no single, clean dashboard where you can compare storms or track regional vulnerability over time. I wanted to build something closer to that — a structured look at the data that shows patterns rather than just events.

How I Put It Together

Getting the data was the hard part. NDRRMC publishes final situation reports as PDFs, not spreadsheets. I manually extracted key figures — affected families, deaths, injured, damaged houses, and peso damage — from these reports. It wasn't glamorous work, but there's no shortcut when the data lives in unstructured documents.

Normalizing damage for inflation was important because a billion pesos in 2014 isn't the same as a billion pesos in 2020. I used PSA inflation rates to convert all damage figures to a common year's peso value, which made year-over-year comparisons honest.

For the regional mapping, I assigned each typhoon's impact to the regions listed in the NDRRMC reports. Some typhoons hit 5 or 6 regions; others were concentrated in 1 or 2. The cumulative view — total damage and total affected people per region across all 23 storms — revealed the geography of vulnerability.

What the Data Showed

Yolanda (Hainan, 2013 — though its aftermath data extended into 2014 in the NDRRMC reports) still cast the longest shadow in the dataset. Its damage figures dwarfed everything else. But what struck me more was how consistently Bicol and Eastern Visayas appeared at the top of the affected-population rankings, storm after storm. These regions aren't just occasionally unlucky — they're structurally exposed.

Agricultural damage often exceeded infrastructure damage. Farms don't have the same recovery capacity as roads or buildings. A road can be repaired in weeks; a destroyed rice crop means an entire season's income is gone. The agricultural losses in the data represent not just economic damage but food security risks that ripple outward for months.

• Bicol and Eastern Visayas were the most frequently and severely affected regions
• Agricultural losses often exceeded infrastructure damage in peso terms
• Yolanda's impact dwarfed all other storms combined
• Even "weaker" typhoons affected hundreds of thousands of people each
• Inflation-adjusted damage showed no clear downward trend over the period

Looking Back

Working with disaster data is heavy. Behind every number is a family that lost a home, a farmer who lost a harvest, a community that had to rebuild from scratch — sometimes for the second or third time. The data doesn't capture resilience, mutual aid, or the quiet ways people recover. It only captures what was lost.

But that's precisely why it matters. If the same regions keep absorbing the worst damage year after year, that's a signal for where disaster preparedness investment should go. The data makes that argument more convincingly than any opinion can.