https://www.ngdc.noaa.gov/mgg/topo/globe.html
https://www.ngdc.noaa.gov/mgg/topo/report/globedocumentationmanual.pdf
Elevation Code: elevations.py
This analysis counts tropical storms and hurricanes and corrects for linear trends in tropical storm and hurricane detection. Of course the detection of tropical storms and hurricanes has not changed linearly, but 100 years of near-land storm detection shows about a 7% increase to a 23% decrease in the percent of storms near land. Using the linear correction is probably adequate for the oldest data and some newer data as satellite-based detection is still improving.
The storms near land (versus storms not near land) should be constant. There's no physical reason for storm tracks to change from nearing land to not nearing land over the past 100 years. But there is a reasonable likelihood that storms near land were easier to detect 100 years ago, whereas storms over the open ocean are now equally easy to detect.
However, the change is different for different storm strengths:
Min speed (knots) | Max speed (knots) | Trend per year |
---|---|---|
35 | 54 | -0.00230 |
55 | 69 | 0.000785 |
70 | 84 | -0.00209 |
85 | 104 | -0.001615 |
105 | 124 | -0.001824 |
125 | and up | -0.001067 |
The counts in each category for the year is adjusted by multiplying by (1 - (trend * (YEARS[1] - current_year)))
The corrected counts show a large increase in the strongest hurricanes (>= 125 knots). There's an increase in stronger tropical storms, some increase in weaker tropical storms. The other hurricane counts are flat.
Code: tropicalnumbers.py | Data: hurdat.csv | Instructions: README
Corrected counts by storm strength: |
---|
|
Detection trends shown by change in percent of storms that travel near land: |
|