Many people are aware of the Enhanced Fujita (EF) Scale for tornadoes. Most also know about the Saffir-Simpson Hurricane Wind Scale. However there is also a rating scale for Winter Storms that fewer people know about. In 2004, Paul Kocin and Louis Uccellini from the National Weather Service (Kocin and Uccellini, 2004) developed the Northeast Snowfall Impact Scale (NESIS). Below is a description from the National Climate Data Center (NCDC) website for NESIS:
“The index differs from other meteorological indices in that it uses population information in addition to meteorological measurements. Thus NESIS gives an indication of a storm’s societal impacts. This scale was developed because of the impact Northeast snowstorms can have on the rest of the country in terms of transportation and economic impact.
NESIS scores are a function of the area affected by the snowstorm, the amount of snow, and the number of people living in the path of the storm. The diagram below illustrates how NESIS values are calculated within a geographical information system (GIS). The aerial distribution of snowfall and population information are combined in an equation that calculates a NESIS score which varies from around one for smaller storms to over ten for extreme storms. The raw score is then converted into one of the five NESIS categories. The largest NESIS values result from storms producing heavy snowfall over large areas that include major metropolitan centers. For details on how NESIS scores are calculated at the National Climatic Data Center, see Squires and Lawrimore (2006).”
So, let’s compare some storms from the past few decades to see how they stack up against each other.
When you compare these two storms, they really don’t look too different at first glance The Baltimore, Washington and Philadelphia Metropolitan Areas seem to have similar snowfall amounts. Even the southern extent of the 10″ snowfall is pretty much the same (Ocean City, MD to Charlottesville, VA). However the key difference for these storms is their impact to the major metropolitan areas further up the I-95 corridor. The 2003 event has 20-30″ snowfall in both the New York City and Boston Metropolitan Areas while essentially no snow fell in the 2010 event.
Those who are focused only on the science of the hazard may say that population shouldn’t factor in to historical comparisons. I would counter that thought with a look at the other hazards to see if this comparison is used elsewhere:
Hurricanes – The Saffir-Simpson Scale is a true wind speed scale which doesn’t incorporate population. However the perception among many is that the Hurricanes that hit populated areas are worse than those that don’t. There are some reasons why this perception exists and I won’t go into that, but this would help to explain why a Category 1 in the Mid-Atlantic and Northeast Corridor seems so much worse than a Category 1 hitting in the Big Bend region or other lesser-populated regions of the Gulf Coast. Conclusion – Safir-Simpson measures raw intensity, but perception may be different
Tornadoes – The EF-Scale is a damage scale, so it needs structures (especially well built structures) to help determine intensity for higher end events. When the structure design is exceeded, it becomes more difficult to rate the tornado as a higher intensity event. For example, mobile home damage shows up to EF-3 damage, but beyond that, mobile homes are. Likewise, well-built structures allow engineers to better determine the true wind speeds (see Joplin Storm Survey from NWS Springfield), and there are just more of these structures in higher density populated areas. (Conclusion – EF Scale measures impact, not just raw wind speeds)
Earthquakes – There are two scales. The first is the instrumental scale called the Richter Scale which measures magnitude. The second is an impact scale called the Modified Mercalli Intensity Scale which measures intensity. While the Richter scale measures the impact. For example, both the Haiti Earthquake of January 2010 and the Christ Church earthquake of September 2010 measured 7.0 on the Richter Scale. However, on the MMI Intensity scale, only 2k people were in MMI 9 or 10 Intensity areas. For Haiti, approximately 1 million people were in MMI 9 or 10 Intensity areas. Conclusion – Earthquakes are measured both ways.
Flooding – To my knowledge, there is no scale for flooding severity, but in general, people compare specific floods to the FEMA Flood Map identified “100 year flood zone”. I won’t get into any issues with using that terminology, but all floods are different. When comparing Mississippi River flooding for 1993 to 2011, there are key similarities and differences between these floods. Having a way to know how bad the flooding is over a large area would be valuable, but would still be difficult to put together. Individual gages can have historical comparisons, but once multiple gages are brought in, it becomes difficult. Conclusion – Flooding could be measured either way, but would need more work.
So, to wrap this up, when we look at other major hazards impacting the US, the is a pattern emerging that several of these hazards are starting to look at impact in addition to the scientific magnitude of the event. As these methods continue to improve, it will help many people better understand “how bad is bad”, and “when did we last see this type of an event”.
Note: The National Climatic Data Center is also producing a new regional snowfall index; the Regional Snowfall Impact Scale (ReSIS) that will look at snow impacts regionally – great topic for a follow-up article!