US Army Corps of Engineers
Engineer Research and Development Center Website

Damaging Ice Storm GIS

Damaging Ice Storm Geographic Information System - click image below for data

Published Aug. 5, 2014

 Click image for data - The storm footprints in the GIS delineate the area where ice sensitive structures were damaged. e.g., overhead power, phone and cable TV lines, communication towers, and trees.
Click image for data - The storm footprints in the GIS delineate the area
where ice sensitive structures were damaged. e.g., overhead power,
phone and cable TV lines, communication towers, and trees.

Damaging freezing rain storms

We have compiled information on damaging freezing rain storms (ice storms, silver thaws, blue northers) from many CRREL projects between 1993 and 2009 to estimate the equivalent radial thickness of ice on ice-sensitive structures for a 50-year mean recurrence interval for the contiguous 48 states, Alaska, and the southern tier of Canada. A 50-year ice map for freezing rain in the United States is in ASCE Standard 7 Minimum Design Loads for Buildings and Other Structures (ASCE 2010). Equivalent radial ice thicknesses from past freezing rain events were estimated by modeling the accretion of ice on wires using hourly weather data and daily, six-hourly, or hourly precipitation data from weather stations across the United States and neighboring Canadian provinces.

We identified storms with significant modeled ice thicknesses and collected information on those events from contemporaneous newspaper reports, Storm Data (NOAA 1949 to present) and its predecessors,  a compilation of communication tower failures (Mulherin 1996), FEMA Mitigation Reports, utility reports, journal articles, and other publications. An extreme value analysis using these modeled ice thicknesses from freezing rain provides equivalent radial ice thickness estimates for mean recurrence intervals between 25 and 500 years for the design of ice sensitive structures.

The period of record of this database is from the late 1940s to the present. The beginning of the period of record was defined by the availability of hourly weather data in electronic format.Electronic records for weather stations in Canada begin in 1953. In the United States, electronic records for some stations begin in the late 1940s but for many others data were not available until the early 1970s. With the low station density prior to that time, it is likely that we have missed damaging ice storms and have not mapped the full extent of other storms. The database has been updated through the spring of 2014 based on information in newspaper articles.

The storm footprints in the GIS delineate the area where ice sensitive structures (overhead power, phone and cable TV lines, communication towers, and trees) were damaged. Ice storms that cause only slippery roads are not mapped. Each storm footprint has an associated description, organized by state. The information provided in the storm descriptions includes, when available, the weather associated with the damage to ice sensitive structures, observations of ice thickness, the cause of the damage, the severity of the damage, including the number of electric utility customers who lost power and the duration of the power outages, and local or national emergency or disaster declarations that ensued. Any associated flooding is mentioned. Difficulties causing delays in repairing infrastructure damage, including reduced access, cold, wind, snow, water, and mud, are described. Descriptions of memorable ice storms often include comparisons to previous ice storms, or tornados and hurricanes. These comparisons are typically attributed to long-time residents, utility spokespeople, public officials, or local weather observers. These sources may not agree on their assessments of "this is the worst storm since...". When contradictory assessments are provided in newspaper articles, Storm Data, or utility reports, they are all included in the storm description.

This database is intended to be restricted to freezing rain storms. However, freezing rain often occurs with snow, so there may be tree damage and associated power outages caused by the weight of accumulated snow and ice. There are also some events in which the overhead lines might have been damaged by in cloud icing or freezing drizzle, with no measured precipitation, rather than by freezing rain. When in cloud icing is suspected, that information is included in the storm description.

Cause of damage

The causes of damage to ice sensitive structures are many and varied.

  • Tree branches are broken by the weight of the ice (gravity load), or by the wind load on the ice-covered tree. Tree trunks break or trees uproot if the wind-on-ice load is too high, with uprooting more likely when the ground is wet rather than frozen. If a storm occurs early enough that the trees still have leaves, then the additional weight and area of the ice-covered leaves adds to the gravity and wind loads. Young, flexible trees tend to bend, sometimes to the ground, while impressively large old, stiff trees, with years of accumulated damage from insects, animals, wind, lightning, and snow and ice loads, are more likely to break.
  • Power distribution lines often fail when ice-covered trees and branches fall on the wires, breaking the wires themselves or overloading the poles that then break or pull out of the ground. Tree damage to power lines can be ameliorated if the utility keeps the nearby trees cut back well away from the wires. However, tree trimming can often be a contentious issue between utilities and local residents and when not done by experts can weaken the tree. And on streets with scenic designations in rural areas, tree trimming may be restricted by regulation. Wires and poles also fail when the combination of the ice and wind load exceed their breaking strength. Failure can occur under relatively low loads because of prior damage in previous storms, or from vandalism, fatigue (e.g. aeolian vibration), or accumulated damage over the years. Power outages in freezing rain storms are sometimes initiated by a car sliding off the slippery road and hitting a pole. Sagging wires in a span are sometimes snagged by a truck driving underneath that then pulls down the poles at the ends of the span. The failure of any pole can initiate a long line of pole failures. Lines that have been repaired may fail again as ice-covered trees and branches continue to fall in cold and sometimes windy weather after freezing rain ends. As the ice begins to melt and fall off the sagging wires, they sometimes spring up, and are damaged as they hit the overhanging branches. This phenomenon is called sleet jumping.
  • Power transmission lines sometimes fail when ice-covered trees fall on them, but those lines tend to be in wide right-of-ways and the conductors are relatively high, so only tall trees near the edge of the right-of-way are potential hazards. Shield wires, conductors, insulator strings, and crossarms fail if the combination of the ice and wind load exceeds their breaking strength. Sometimes a broken shield wire falls across a conductor, causing the conductor to burn through. Shield wire strength may be reduced by lightning strikes burning through strands of the wire. The wire system and support structures may be damaged by galloping and fail under relatively low ice loads. Galloping is a low frequency, high amplitude oscillation of the cable and may affect only a few spans. It occurs when the accreted ice shape leads to a lift force on the wire from the wind blowing by it. Galloping often occurs in relatively windy conditions with relatively small ice accretions and can persist for hours or even days. The failure of components in the wire system can initiate a cascade failure in which the unbalanced loading on a support structure (e.g. pole, H-frame, lattice tower) causes it to collapse. Cascade failures sometimes propagate for tens of miles.
  • Phone and cable TV service is sometimes lost when overhead lines are damaged. However, these lines can still be used when they are on the ground if they remain intact, so these outages typically affect only a relatively small number of customers, compared to those with no power. Cellular phone service may also be affected if there is no power to some or all of the cell phone towers in a region and tower backup generators are not operating.
  • Damaged telegraph lines and long distance phone lines were the main story in ice storms for many years, with lists of towns that were cut off from outside communication. These lines often had many wires on each of a number of crossarms, so the weight of ice on the poles and the added wind drag could be substantial with even relatively small accumulations of ice.
  • Television, radio, and cell phone towers are damaged in some freezing rain storms. Ice on the components of lattice towers and on guy wires adds to the vertical load on the structure and increases the wind drag. Sometimes one tower in a group of towers fails, taking other towers with it. Sometimes ice falling from towers damages the buildings below.  

Reported ice thicknesses and equivalent radial ice thickness

Ice thicknesses in the storm summaries are from newspaper articles, NOAA's Storm Data publication, electric utility press releases, and FEMA mitigation reports. Unless stated otherwise, if the thickness reported is on a wire or branch, it should be taken as either the maximum dimension of the ice accretion, including the wire or branch diameter, or the maximum radial thickness of the ice. The thickness of ice on a horizontal surface (road, sidewalk, picnic table) or vertical surface (side of a building, leg of a tower) may also be reported.

The equivalent radial ice thickness Req is the thickness that ice on a cylinder would have if it were distributed to a uniform thickness around the cylinder (Jones 1998). In the same weather conditions the amount of ice as measured by the equivalent radial ice thickness is independent of cylinder diameter if the cylinder orientations to horizontal and to the wind direction are the same. Req can be calculated from a sample of ice using the ice mass m, the length of the sample L, and the diameter of the cylinder d:


Examples of some of the variety of shapes of ice accretions on trees, wires, and cylinders are shown in the following photos taken by members of the CRREL Ice Storm Team in freezing rain storms across the country.[R4] 

Conductor Galloping

This Bonneville Power Administration video of conductor gallop was shot by a line crew in eastern Idaho Dec. 26, 1998. It’s estimated that there was a thin accumulation of glaze ice on the wires with a 15-25 mph wind from the south. The conductor motion caused a series of outages Dec. 25 and 26.


Most of the ice storm photos on this web page were taken by the members of CRREL's Ice Storm Team, Kerry Claffey, Nate Mulherin, and Kathy Jones. Credit for photos provided by other photographers is included in the photo caption.

This project was conducted for the American Lifelines Alliance of the Multihazard Mitigation Council with funding provided under a contract between the National Institute of Building Sciences and the Federal Emergency Management Agency of the Department of Homeland Security.

The projects, for which the ice storm damage information was originally compiled were funded by: American Lifelines Alliance; Bonneville Power Administration; CEATI's Overhead Line Design Issues & Wind and Ice Storm Mitigation Group; Cold Regions Research and Engineering Laboratory, U.S. Army Corps of Engineers; Duke Energy; Electric Power Research Institute; Entergy; Florida Power and Light; Mid American Energy; New Hampshire PUC; New York Power Authority; Newfoundland and Labrador Hydro; Northern States Power; Power Engineers; Progress Energy; Santee Cooper; South Carolina Electric and Gas/SCANA; Southern Company; Tennessee Valley Authority; Vermont Electric Power Company; Western Area Power Administration.


General references on freezing rain storms are below. References for particular storms are included with the storm descriptions.

  • ASCE 7 Ice Load Task Committee (2010) Minimum design loads for buildings and other structures ASCE 7-10, Chapter 10 Ice Loads–Atmospheric Icing, American Society of Civil Engineers, Reston, Virginia.
  • Bernstein, B.C. and B.G. Brown (1997) A climatology of supercooled large drop conditions based on surface observations and pilot reports of icing, paper 4.3, The 7th Conference on Aviation, Range and Atmospheric Meteorology, Long Beach, California, 2-7 Feb 1997.
  • Davis, R.E. and D.A. Gay (1993) Freezing rain and sleet climatology of the Southeastern U.S.A., Research Paper #052593, Southeast Regional Climate Center, South Carolina Water Resources Commission, Columbia, South Carolina.
  • Jones, K.F. (1998) A simple model for freezing rain ice loads, Atmospheric Research, 46, 87-97.
  • Jones, K.F. (1999) Ice storms, trees and power lines, Proceedings of the 10th International Conference on Cold Regions Engineering, August 1999, Lincoln, New Hampshire, pp 757-767.
  • Jones, K.F., R. Thorkildson and J.N. Lott (2002) The development of the map of extreme ice loads for ASCE Manual 74, Electrical Transmission in a New Age, Omaha, ASCE, Reston Virginia, pp 9-31. Also published as The Development of a U.S. Climatology of Extreme Ice Loads. [R10] 
  • Klaassen, J, S Cheng, H Auld, Q Li, E Ros, Geast M, Li G and R Lee (2003) Estimation of Severe Ice Storm Risks for South-Central Canada. Meteorological Service of Canada - Ontario Region, Environment Canada. Study prepared for the Office of Critical Infrastructure Protection and Emergency Preparedness, Ottawa. 2003 Minister of Public Works and Government Services Catalogue No. PS4-6/2004E-PDF ISBN: 0-662-37712-5.
  • McKay, G.A. and H.A. Thomson (1969) Estimating the hazard of ice accretion in Canada from climatological data, Journal of Applied Meteorology, 8, pp 927-935.
  • Mulherin, N.D.(1996) Atmospheric Icing and Tower Collapse in the United States, Proceedings of the 7th International Workshop on Atmospheric Icing of Structures, Chicoutimi, Quebec, Canada, June 1996, p 457.
  • National Oceanic and Atmospheric Administration, "Storm Events Database"
  • Robbins, C.C. and J.V. Cortinas (1996) A climatology of freezing rain in the contiguous United States: preliminary results, 15th Conference on Weather and Forecasting, Norfolk, Virginia, 19-23 August, American Meteorological Society, pp 124-126.
  • Simpson, P.O. (1999) Tree damage to electric utility infrastructure, assessing and managing the risk from storms, Proceedings of the 10th International Conference on Cold Regions Engineering, August 1999, Lincoln, New Hampshire, pp 768-778.
  • USFS (1994) Forest Service Handbook FSH6609.14 Telecommunications Handbook, R3 Supplement 6609.14-94-2 Effective 5/2/94, United States Forest Service, Washington, D.C.
  • Vilcans, J. and D. Burnham (1989) Climatological study to determine the impact of icing on the low level windshear alert system, Volume I Analysis and Volume II Statistics, DOT-TSC-FAA-89-2 and DOT-TSC-FAA-89-3, U.S. Department of Transportation, Transportation Systems Center, Cambridge, Massachusetts.

Work with Us

  • Ice storm photos

    • If you'd like your photos to be considered for an existing storm description - please email them with:
      • the storm identification number
      • descriptive captions and photo credit information
  • Freezing rain storm information

    • If you have new information to be considered for inclusion within in the database - please email with:
      • storm dates
      • storm description

Contact, 603-646-4417

Terrestrial & Cryospheric Sciences Branch (CEERD-RR-G)
US Army Engineer Research and Development Center - Cold Regions Research and Engineering Laboratory