DECEMBER 1998 NOTES AND CORRESPONDENCE 1187

NOTES AND CORRESPONDENCE

The Steadman Chill: An Improvement over Present Scales

ROBERT G. QUAYLE National Climatic Data Center, Asheville, North Carolina

ROBERT G. STEADMAN School of Agricultural Science, La Trobe University, Bundoora, Victoria, Australia 19 February 1998 and 24 June 1998

ABSTRACT Because of shortcomings in the current wind chill formulation, which did not consider the metabolic heat generation of the human body, a new formula is proposed for operational implementation. This formula, referred to as the Steadman wind chill, is based on peer-reviewed research including a heat generation and exchange model of an appropriately dressed person for a range of low and wind speeds. The Steadman wind chill produces more realistic wind chill equivalents than the current NWS formulation. It is easy to determine from tables (calculated by application of a quadratic ®t in both U.S. and metric units) with values accurate to within 1ЊC.

1. Introduction freezing water to actually freeze under various wind and conditions rather than any human physio- Many scientists have noted de®ciencies in the current logical model of heat loss and gain under equivalent wind chill scale (e.g., Court 1992; Dixon and Prior 1987; conditions. The major consequence is that water, not Driscoll 1985, 1992, 1994; Kessler 1993, 1994, 1995; having a metabolic heat source and not being appro- Horstmeyer 1994; Osczevski 1995; Schwerdt 1995; priately attired in human clothing, produces colder wind Steadman 1995). There was a session partly devoted to chills (for given wind and temperature conditions) than this topic at the 22d Annual Workshop on Hazards Re- those experienced by live, clothed humans. Further- search and Applications (13±16 July 1997) sponsored more, the traditional wind chill adaptation makes two by the Natural Hazards Research and Applications In- other assumptions that are not entirely correct. formation Center, Boulder, Colorado. Unquestionably the current wind chill scale has served us well in high- 1) The base wind speed at which the wind chill equiv- lighting the need for extra clothing and protection in alent temperature equals the ambient air temperature cold, windy conditions. But it may have also made us is shifted from 0 to 4 mph, which implies that complacent by exaggerating the chilling effect of the less than 5 mph have a warming effect at a given wind at very cold temperatures; lulling us, by experi- temperature. ence, into the notion that temperatures well below 2) Winds above 40 mph are assumed to have no ad- Ϫ35ЊC(Ϫ31ЊF) are safe, when in fact, they can be ditional chilling effect. dangerous. Pioneering work by Steadman (1971) and his sub- The wind chill concept was pioneered by Siple and sequent re®nements (1979a,b, 1984, 1994; and 1996, Passel (1945), extrapolated for wind speeds above 12 Ϫ1 1997, unpublished manuscripts) offer an improved wind ms , and promulgated by the National Oceanic and chill scale based on peer-reviewed research employing Atmospheric Administration (NOAA) and the media a biophysical energy balance model of the human body. (NWS 1992a,b; Kuhl 1992). The original wind chill is The Steadman model of heat balance of the human body, based on the length of time it took a vessel of near- originally conceived as a tool for textiles and clothing science, explicitly and implicitly includes all the major heat sources and sinks when estimating heat loss or gain to ascertain the comfort level of a person. Figure 1 sche- Corresponding author address: Dr. Robert G. Quayle, Chief, Glob- al Lab., National Climatic Data Center, DOC/NOAA/NES- matically highlights the various heat exchange mecha- DIS, 151 Patton Avenue, Asheville, NC 28801-5001. nisms [German Klima±Michel Model (WMO 1997)] E-mail: [email protected] that are representative of the Steadman approach. While

᭧ 1998 American Meteorological Society

Unauthenticated | Downloaded 10/03/21 04:00 PM UTC 1188 AND FORECASTING VOLUME 13

FIG. 1. The heat balance of a human being: M ϭ metabolic heat production, QH ϭ sensible heat ¯ux, QSW ϭ latent heat ¯ux, QL FIG. 2. Steadman wind chill vs air temperature and wind speed ϭ moisture heat ¯ux, QRe ϭ respiration heat ¯ux (sensible and from the NCDC model. latent). Radiation budget (including extra radiation received from sources external to the body): I ϭ direct solar radiation, D ϭ diffuse solar radiation, R ϭ re¯ected radiation, A ϭ atmospheric reradiation, E ϭ radiation from the surroundings, EKM ϭ radiation from human surfaces (after WMO 1997). 1 (Steadman 1971, 1984, 1994, 1995; and 1996, 1997, unpublished manuscripts). Although apparent temper- ature is robust with respect to most physiological and clothing variables, such as body temperature, moisture similar, the Klima±Michel model was not used in this permeability of clothing, skin resistances, etc., it is sen- study. The Steadman apparent temperature is operative sitive to the extent to which wind penetration lowers for both heat stress and cold stress (called wind chill). the clothing resistances to heat and moisture transfer. Its input variables include air temperature, , This sensitivity derives partly from the fact that the wind speed, and external solar radiation loading, and it clothing resistance predominates, relative to skin and can be applied to various modeled levels of clothing. boundary layer, in cold conditions. The effects of hu- The National Weather Service (NWS 1992a) employs midity are relatively small at low temperatures and can a version of the Steadman apparent temperature scale be neglected. The warming effect of full midday sun as its , using temperature and humidity input can be highly signi®cant but is not considered here, as only, and assumes appropriate clothing and a very light we are addressing only cloudy or shaded conditions and breeze. At cold temperatures the Steadman apparent are considering wind speed and temperature only. temperature scale, used to update the present wind chill The Steadman wind chill matrix from Table 1 (not formulation, offers a replacement for the wind chill for- the long series of Steadman equations) is converted from mula, tables, and graphs used by the NWS and others. metric to U.S. units. Then both metric and U.S. wind The revised scale is simple and provides accurate equa- chill tabulations are ®tted to regression surfaces using tions for both metric and U.S. units. The resulting Stead- second-order multiple-regression analysis. The equa- man wind chill includes heat gained by metabolism; heat tions for these smooth three-dimensional surfaces are retained by clothing; and heat lost by respiration and then used as the Steadman wind chill equations. The conduction, as exacerbated by low temperatures and the regression equations specify the wind chill as the de- wind. For a given temperature and wind speed, it pro- pendent variable, and air temperature and wind speed vides the at which net heat loss as the independent variables. The resulting surfaces ®t would be the same as actual conditions but with calm the base data with an rmse of 0.31ЊC and 0.55ЊF, re- winds. spectively, and an r 2 of 0.9999. The T 2 term in both equations is negligible and is omitted. The resulting equations are as follows. 2. Wind chill calculations The Steadman wind chill equation in metric units is As our basic starting point we employ the complete wind chill ϭ 1.41 Ϫ 1.162V ϩ 0.980T Steadman tables including marginal notations for the SC effects of humidity and extra radiation as shown in Table ϩ 0.0124V 2 ϩ 0.0185(VT), (1)

Unauthenticated | Downloaded 10/03/21 04:00 PM UTC DECEMBER 1998 NOTES AND CORRESPONDENCE 1189 C) Њ 0.491 0.423 0.369 0.310 0.260 0.221 0.188 0.159 0.134 0.112 0.097 Additive humidity effect* ( 7.3 4.6 1.2 1.4 4.0 6.6 9.2 19.8 22.6 25.2 27.9 30.6 33.2 35.9 38.6 41.2 43.8 46.4 49.2 51.8 54.4 57.2 59.8 62.4 65.2 67.9 70.6 11.9 14.4 17.2 10.0 44.2 37.7 30.3 23.4 20.8 18.2 15.4 12.7 Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ 9.6 6.9 3.9 1.4 1.1 3.6 6.0 8.4 15.9 18.4 20.9 23.4 25.9 28.4 30.8 33.3 35.8 38.2 40.7 43.2 45.7 48.2 50.6 53.1 55.6 58.1 60.6 63.1 10.9 13.4 12.1 44.3 38.1 31.4 24.8 22.3 19.8 17.2 14.6 Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ 8.4 5.6 3.2 0.9 1.6 3.9 6.3 8.6 13.4 15.8 18.2 20.6 23.0 25.3 27.7 30.0 32.4 34.8 37.1 39.6 41.9 44.2 46.6 49.0 51.3 53.7 56.1 58.4 13.4 11.0 11.0 44.2 38.4 32.0 25.7 23.3 20.9 18.4 15.9 C, base vapor ) Ϫ Ϫ Ϫ Ϫ Њ Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ ) 1 9.7 7.3 4.6 2.2 0.0 2.4 4.7 7.0 9.3 Ϫ 11.7 14.0 16.3 18.6 20.9 23.2 25.6 27.8 30.1 32.4 34.6 36.9 39.2 41.6 43.8 46.1 48.4 50.7 53.0 55.3 14.4 12.0 44.4 38.6 32.4 26.3 24.0 21.6 19.2 16.8 Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ of body surface at 0 2 Ϫ 9.8 8.4 6.0 3.7 1.4 0.8 3.0 5.3 7.6 12.0 14.3 16.6 18.8 21.0 23.2 25.4 27.6 29.8 32.0 34.3 36.4 38.7 40.9 43.1 45.3 47.6 49.8 52.0 15.4 13.1 10.8 44.4 38.8 32.9 27.0 24.7 22.4 20.1 17.7 Ϫ Ϫ Ϫ Ϫ Ϫ Wind speed (m s Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ 7.7 9.9 7.6 5.4 3.2 1.0 1.1 3.3 5.6 12.1 14.3 16.4 18.6 20.7 22.9 25.0 27.1 29.3 31.4 33.6 35.7 37.8 40.0 42.1 44.2 46.4 48.6 16.6 14.4 12.2 10.0 44.4 39.0 33.4 27.8 25.6 23.4 21.1 18.8 Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ C and colder. Additionally, prolonged exposure of skin to any freezing temperatures can cause injury. Њ 35 Ϫ 5.3 7.4 9.6 9.4 7.3 5.2 3.1 1.0 1.1 3.2 11.7 13.8 15.8 17.9 20.0 22.0 24.1 26.1 28.2 30.3 32.3 34.4 36.4 38.6 40.6 42.7 44.8 18.0 15.9 13.8 11.6 44.6 39.3 34.0 28.7 26.6 24.4 22.3 20.1 Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ Additive effect of extra radiation** (100 W m 2.9 5.0 7.0 9.0 5.6 5.1 4.9 4.1 3.3 3.1 2.8 2.5 9.3 7.3 5.3 3.3 1.2 0..8 ): Wind effects at low temperatures including humidity and added radiation effects. These are the input data used to derive the 11.0 13.1 15.1 17.1 19.1 21.1 23.1 25.1 27.1 29.1 31.1 33.1 35.1 37.1 39.2 41.2 44.8 39.8 34.7 29.6 27.6 25.6 23.6 21.6 19.4 17.4 15.4 13.4 11.4 1 2.0 4.0 6.0 8.0 10.0 12.0 15.0 20.0 Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ C,ms Њ C) Њ 2.0 4.0 6.0 8.0 8.0 6.0 4.0 2.0 0.0 10.0 12.0 14.0 16.0 18.0 20.0 22.0 24.0 26.0 28.0 30.0 32.0 34.0 36.0 38.0 40.0 45.0 40.0 35.0 30.0 28.0 26.0 24.0 22.0 20.0 18.0 16.0 14.0 12.0 10.0 Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ 18 23 30 39 44 49 52 56 60 64 68 73 76 81 86 90 93 96 98 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 VP % Dry bulb ( Approx RH at base 5.18 4.37 3.69 3.10 2.60 2.18 1.81 1.51 1.25 1.03 0.85 0.70 0.58 0.47 0.38 0.31 0.25 0.20 0.16 0.13 9.20 8.40 7.60 6.80 6.00 17.50 17.26 16.98 16.71 16.60 16.40 15.60 14.80 14.00 13.20 12.40 11.60 10.80 10.00 1. Steadman (1996) apparent temperature ( 5.18 4.37 3.69 3.10 2.60 2.18 1.81 1.51 1.25 1.03 0.85 0.70 0.58 0.47 0.38 0.31 0.25 0.20 0.16 0.13 9.35 8.13 7.06 6.11 95.82 73.77 56.22 42.44 37.80 33.62 29.84 26.44 23.38 20.62 18.18 16.00 14.13 12.28 10.73 * Effect of raising relative humidity by 10%, wind absent. Steadman wind chill equations. Danger: Bare skin freezes at wind chill of ABLE saturation Base ** Extra radiation received from sources outside the body. sure (hPa) Vapor pres- T Metric units

Unauthenticated | Downloaded 10/03/21 04:00 PM UTC 1190 WEATHER AND FORECASTING VOLUME 13

TABLE 2. Steadman wind chill (ЊC,msϪ1) from Eq. (1). Danger: Bare skin freezes at wind chill of Ϫ35ЊC and colder. Additionally, prolonged exposure of skin to any freezing temperatures can cause injury. Metric units Additive Wind speed (m sϪ1) humidity Dry bulb (ЊC) 2.0 4.0 6.0 8.0 10.0 12.0 15.0 20.0 effect* (ЊC) 45.0 44.9 44.4 44.0 43.7 43.5 43.4 43.4 43.9 40.0 39.8 39.1 38.5 38.0 37.7 37.4 37.1 37.2 35.0 34.7 33.9 33.1 32.4 31.8 31.3 30.8 30.4 30.0 29.6 28.6 27.6 26.8 26.0 25.3 24.5 23.7 28.0 27.6 26.5 25.4 24.5 23.7 22.9 22.0 21.0 26.0 25.6 24.4 23.3 22.3 21.3 20.5 19.5 18.3 24.0 23.5 22.3 21.1 20.0 19.0 18.1 17.0 15.6 22.0 21.5 20.2 18.9 17.7 16.7 15.7 14.5 12.9 20.0 19.5 18.0 16.7 15.5 14.4 13.3 12.0 10.2 0.491 18.0 17.4 15.9 14.5 13.2 12.0 10.9 9.4 7.5 0.423 16.0 15.4 13.8 12.4 11.0 9.7 8.5 6.9 4.8 0.369 14.0 13.4 11.7 10.2 8.7 7.4 6.1 4.4 2.1 0.310 12.0 11.3 9.6 8.0 6.5 5.0 3.7 1.9 Ϫ0.6 0.260 10.0 9.3 7.5 5.8 4.2 2.7 1.3 Ϫ0.6 Ϫ3.3 0.221 8.0 7.3 5.4 3.6 1.9 0.4 Ϫ1.1 Ϫ3.1 Ϫ6.0 0.188 6.0 5.2 3.3 1.4 Ϫ0.3 Ϫ2.0 Ϫ3.5 Ϫ5.7 Ϫ8.7 0.159 4.0 3.2 1.2 Ϫ0.7 Ϫ2.6 Ϫ4.3 Ϫ5.9 Ϫ8.2 Ϫ11.4 0.134 2.0 1.2 Ϫ0.9 Ϫ2.9 Ϫ4.8 Ϫ6.6 Ϫ8.3 Ϫ10.7 Ϫ14.1 0.112 0.0 Ϫ0.9 Ϫ3.0 Ϫ5.1 Ϫ7.1 Ϫ9.0 Ϫ10.7 Ϫ13.2 Ϫ16.8 0.097 Ϫ2.0 Ϫ2.9 Ϫ5.1 Ϫ7.3 Ϫ9.3 Ϫ11.3 Ϫ13.1 Ϫ15.7 Ϫ19.5 Ϫ4.0 Ϫ4.9 Ϫ7.2 Ϫ9.5 Ϫ11.6 Ϫ13.6 Ϫ15.5 Ϫ18.2 Ϫ22.2 Ϫ6.0 Ϫ7.0 Ϫ9.4 Ϫ11.6 Ϫ13.8 Ϫ15.9 Ϫ17.9 Ϫ20.7 Ϫ24.9 Ϫ8.0 Ϫ9.0 Ϫ11.5 Ϫ13.8 Ϫ16.1 Ϫ18.3 Ϫ20.3 Ϫ23.3 Ϫ27.6 Ϫ10.0 Ϫ11.0 Ϫ13.6 Ϫ16.0 Ϫ18.4 Ϫ20.6 Ϫ22.7 Ϫ25.8 Ϫ30.3 Ϫ12.0 Ϫ13.1 Ϫ15.7 Ϫ18.2 Ϫ20.6 Ϫ22.9 Ϫ25.1 Ϫ28.3 Ϫ33.0 Ϫ14.0 Ϫ15.1 Ϫ17.8 Ϫ20.4 Ϫ22.9 Ϫ25.3 Ϫ27.6 Ϫ30.8 Ϫ35.7 Ϫ16.0 Ϫ17.1 Ϫ19.9 Ϫ22.6 Ϫ25.1 Ϫ27.6 Ϫ30.0 Ϫ33.3 Ϫ38.4 Ϫ18.0 Ϫ19.2 Ϫ22.0 Ϫ24.7 Ϫ27.4 Ϫ29.9 Ϫ32.4 Ϫ35.8 Ϫ41.1 Ϫ20.0 Ϫ21.2 Ϫ24.1 Ϫ26.9 Ϫ29.6 Ϫ32.3 Ϫ34.8 Ϫ38.4 Ϫ43.8 Ϫ22.0 Ϫ23.2 Ϫ26.2 Ϫ29.1 Ϫ31.9 Ϫ34.6 Ϫ37.2 Ϫ40.9 Ϫ46.5 Ϫ24.0 Ϫ25.3 Ϫ28.3 Ϫ31.3 Ϫ34.1 Ϫ36.9 Ϫ39.6 Ϫ43.4 Ϫ49.2 Ϫ26.0 Ϫ27.3 Ϫ30.4 Ϫ33.5 Ϫ36.4 Ϫ39.2 Ϫ42.0 Ϫ45.9 Ϫ51.9 Ϫ28.0 Ϫ29.3 Ϫ32.5 Ϫ35.7 Ϫ38.7 Ϫ41.6 Ϫ44.4 Ϫ48.4 Ϫ54.6 Ϫ30.0 Ϫ31.4 Ϫ34.7 Ϫ37.8 Ϫ40.9 Ϫ43.9 Ϫ46.8 Ϫ50.9 Ϫ57.3 Ϫ32.0 Ϫ33.4 Ϫ36.8 Ϫ40.0 Ϫ43.2 Ϫ46.2 Ϫ49.2 Ϫ53.4 Ϫ60.0 Ϫ34.0 Ϫ35.4 Ϫ38.9 Ϫ42.2 Ϫ45.4 Ϫ48.6 Ϫ51.6 Ϫ56.0 Ϫ62.7 Ϫ36.0 Ϫ37.5 Ϫ41.0 Ϫ44.4 Ϫ47.7 Ϫ50.9 Ϫ54.0 Ϫ58.5 Ϫ65.4 Ϫ38.0 Ϫ39.5 Ϫ43.1 Ϫ46.6 Ϫ49.9 Ϫ53.2 Ϫ56.4 Ϫ61.0 Ϫ68.1 Ϫ40.0 Ϫ41.5 Ϫ45.2 Ϫ48.7 Ϫ52.2 Ϫ55.6 Ϫ58.8 Ϫ63.5 Ϫ70.8 Additive effects of extra radiation** (100 W mϪ2 of body surface at 0ЊC, base ) 5.6 5.1 4.9 4.1 3.3 3.1 2.8 2.5

* Effect of raising relative humidity by 10%, wind absent. ** Extra radiation received from sources outside the body.

where wind chill is expressed in degrees Celsius, V is is the wind speed in statute miles per hour, and T is the wind speed in meters per second, and T is air temper- air temperature in degrees Fahrenheit. ature in degrees Celsius. A three-dimensional plot of the Steadman wind chill The Steadman wind chill equation in U.S. units is surface as a function of temperature and wind speed is shown in Fig. 2, and the resulting table, rounded to wind chillSF ϭ 3.16 Ϫ 1.20V ϩ 0.980T tenths of degrees, is shown in Table 2. The rounded ϩ 0.0044V 2 ϩ 0.0083(VT), (2) difference between the complex Steadman (1996, un- published manuscript) apparent temperature formulation where wind chill is expressed in degrees Fahrenheit, V (Table 1) and this Steadman wind chill (Table 2) is

Unauthenticated | Downloaded 10/03/21 04:00 PM UTC DECEMBER 1998 NOTES AND CORRESPONDENCE 1191 shown in Table 3. The differences between the two for- TABLE 3. Difference (ЊC, rounded): Steadman wind chill from Table mulations do not exceed 1ЊC. 2 minus Steadman (1996) apparent temperature from Table 1. For domestic U.S. application outside the scienti®c Metric units community, the Steadman wind chill from Eq. (2) above Wind speed (m sϪ1) is applied to the marginal wind speed (statute miles per Dry bulb hour) and air temperature (degrees Fahrenheit) values (ЊC) 2.0 4.0 6.0 8.0 10.0 12.0 15.0 20.0 from NWS (1992b) (also shown in Table 5). The results, 45.0 0 Ϫ0 Ϫ0 Ϫ1 Ϫ1 Ϫ1 Ϫ1 Ϫ0 rounded to whole degrees Fahrenheit, are shown in Ta- 40.0 0 Ϫ0 Ϫ0 Ϫ1 Ϫ1 Ϫ1 Ϫ1 Ϫ1 ble 4. We have carried the marginal notations regarding 35.0 0 Ϫ0 Ϫ0 Ϫ0 Ϫ1 Ϫ1 Ϫ1 0 Ϫ2 30.0 0 Ϫ0 Ϫ0 Ϫ0 Ϫ0 Ϫ0 Ϫ0 0 the warming effect of 100 W m in degrees Fahrenheit, 28.0 0 Ϫ0 Ϫ0 Ϫ0 Ϫ0 Ϫ0 Ϫ0 0 interpolated from a conversion of the same entries in 26.0 Ϫ0 Ϫ0 Ϫ0 Ϫ0 Ϫ0 Ϫ0 Ϫ0 0 Table 1. 24.0 Ϫ0 Ϫ0 Ϫ0 Ϫ0 Ϫ0 Ϫ0 Ϫ0 0 22.0 Ϫ0 0 0 0 Ϫ0 Ϫ0 Ϫ0 0 20.0 0 0 0 0 Ϫ0 Ϫ0 Ϫ0 0 3. Comparisons and recommendations 18.0 0 0 0 0 0 Ϫ0 Ϫ0 0 16.0 0 0 0 0 Ϫ0 0 0 0 For comparison, Table 5 shows the current, highly 14.0 Ϫ0 0 0 0 0 1 1 1 popular NWS wind chill (NWS 1992b) as derived from 12.0 Ϫ0 0 0 0 0 0 0 1 the following: 10.0 0 0 0 1 1 0 0 1 8.0 Ϫ0 0 0 1 0 0 0 1 0.5 wind chillNWS ϭ 0.0817(3.71V ϩ 5.81 Ϫ 0.25V) 6.0 Ϫ0 0 0 0 0 0 0 0 4.0 Ϫ0 0 0 0 0 0 0 0 ϫ (T Ϫ 91.4) ϩ 91.4, (3) 2.0 Ϫ0 0 0 0 0 0 0 0 0.0 Ϫ0 0 0 1 0 0 0 0 where V is the wind speed in statute miles per hour and Ϫ2.0 0 0 0 0 0 0 0 0 T is the air temperature in degrees Fahrenheit. Ϫ4.0 0 0 0 0 0 0 0 0 Table 6 displays the difference between the current Ϫ6.0 0 0 0 0 0 0 0 0 NWS wind chill from Eq. (3), Table 5, and the Steadman Ϫ8.0 0 0 0 0 0 0 0 0 wind chill from Eq. (2), Table 4. The errors in the current Ϫ10.0 Ϫ0 0 0 0 0 0 0 0 wind chill can be quite large, with the current scale being Ϫ12.0 0 0 0 0 0 0 0 0 Ϫ14.0 0 0 0 0 0 0 Ϫ0 0 as much as 33ЊF too cold at a temperature of Ϫ45ЊF Ϫ16.0 Ϫ0 0 0 0 0 0 Ϫ0 0 and a wind speed of 25 mph. These errors could be Ϫ18.0 Ϫ0 Ϫ0 0 0 0 0 Ϫ0 0 dangerous by promoting the notion (gained through ex- Ϫ20.0 Ϫ0 Ϫ0 0 0 0 0 Ϫ0 Ϫ0 perience) that dangerously cold temperatures can be eas- Ϫ22.0 Ϫ0 Ϫ0 0 0 0 Ϫ0 Ϫ0 Ϫ0 ily tolerated. For instance, assume that a person uses Ϫ24.0 Ϫ0 Ϫ0 0 0 Ϫ0 0 Ϫ0 Ϫ0 the current NWS scale at a temperature of Ϫ5ЊF and a Ϫ26.0 Ϫ0 Ϫ0 0 Ϫ0 Ϫ0 Ϫ0 Ϫ0 Ϫ0 Ϫ28.0 Ϫ0 Ϫ0 0 0 0 Ϫ0 Ϫ0 Ϫ0 wind speed of 20 mph, and calculates a wind chill of Ϫ30.0 Ϫ0 Ϫ0 Ϫ0 Ϫ0 Ϫ0 Ϫ0 Ϫ0 Ϫ0 Ϫ46ЊF (from Table 5). The person does not realize that Ϫ32.0 Ϫ0 Ϫ0 Ϫ0 Ϫ0 Ϫ0 Ϫ0 Ϫ0 Ϫ0 a better estimate is Ϫ25ЊF (from Table 4). Thus the Ϫ34.0 Ϫ0 Ϫ0 Ϫ0 Ϫ0 Ϫ0 Ϫ0 Ϫ0 Ϫ0 person may believe that a wind chill of Ϫ46ЊF can be Ϫ36.0 Ϫ0 Ϫ0 Ϫ0 Ϫ0 Ϫ0 Ϫ0 Ϫ0 Ϫ0 easily tolerated, when in fact, the experience should Ϫ38.0 Ϫ0 Ϫ0 Ϫ0 Ϫ0 Ϫ0 Ϫ0 Ϫ0 Ϫ0 actually relate to Ϫ25ЊF. Suppose that later the same Ϫ40.0 Ϫ0 Ϫ0 Ϫ0 Ϫ0 Ϫ0 Ϫ0 Ϫ0 Ϫ0 individual experiences a temperature of Ϫ35ЊF and a wind speed of 5 mph. The current NWS wind chill (Table 5) will give a reading of Ϫ41ЊF, a bit warmer than his earlier experience. Since he believes he toler- way to help the public differentiate between the old and ated a wind chill of Ϫ46ЊF fairly well (when in actuality new sales is to name the new scale the Steadman wind he was experiencing Ϫ25ЊF), he may now feel safe at chill or the ``New Wind Chill.'' We could also note that a wind chill of Ϫ41ЊF. However, the correct Steadman the old tables were too cold and had many entries that wind chill for a temperature of Ϫ35ЊF and a wind speed reached below Ϫ100ЊF, whereas the new scale has only of 5 mph from Table 4 is Ϫ38ЊF, a rather dangerous one entry that cold. This would make the old and new level. scales easy to distinguish from each other. Or, we could Taking into consideration the the body of literature try to introduce the term ``apparent temperature'' into on apparent temperature and wind chill, and the possible the public lexicon to replace the old ``wind chill,'' and consequences of continued use of the current NWS wind perhaps also ``heat index.'' Our work is cut out for us. chill scale, the logical thing to do is adopt the Steadman wind chill. This will not be easy. The old wind chill Acknowledgments. We offer special thanks to Dr. Lar- scale is so entrenched and has been reproduced in so ry Kalkstein for his suggestions and Dr. Catherine Fel- many popular publications, brochures, and the like that ton for an extraordinarily helpful and thorough review a major public education campaign will be required. One of this material.

Unauthenticated | Downloaded 10/03/21 04:00 PM UTC 1192 WEATHER AND FORECASTING VOLUME 13

TABLE 4. Steadman wind chill (ЊF, mph) from Eq. (2). Danger: Bare skin freezes at wind chill of Ϫ31ЊF and colder. Additionally, prolonged exposure of skin to any freezing temperatures can cause injury. U.S. units Wind speed (mph) Dry bulb (ЊF) 4 5 10 15 20 25 30 35 40 45 35 34 33 29 25 21 17 14 11 8 5 30 29 28 24 19 15 12 8 5 2 Ϫ1 25 24 23 18 14 10 6 2 Ϫ2 Ϫ5 Ϫ8 20 19 18 13 8 4 Ϫ0 Ϫ4 Ϫ8 Ϫ12 Ϫ15 15 14 13 8 3 Ϫ2 Ϫ6 Ϫ10 Ϫ14 Ϫ18 Ϫ22 10 9 7 2 Ϫ3 Ϫ8 Ϫ12 Ϫ17 Ϫ21 Ϫ25 Ϫ28 5 4 2 Ϫ3 Ϫ8 Ϫ13 Ϫ18 Ϫ23 Ϫ27 Ϫ31 Ϫ35 0 Ϫ2 Ϫ3 Ϫ8 Ϫ14 Ϫ19 Ϫ24 Ϫ29 Ϫ33 Ϫ38 Ϫ42 Ϫ5 Ϫ7 Ϫ8 Ϫ14 Ϫ19 Ϫ25 Ϫ30 Ϫ35 Ϫ40 Ϫ44 Ϫ49 Ϫ10 Ϫ12 Ϫ13 Ϫ19 Ϫ25 Ϫ31 Ϫ36 Ϫ41 Ϫ46 Ϫ51 Ϫ55 Ϫ15 Ϫ17 Ϫ18 Ϫ24 Ϫ30 Ϫ36 Ϫ42 Ϫ47 Ϫ52 Ϫ57 Ϫ62 Ϫ20 Ϫ22 Ϫ23 Ϫ30 Ϫ36 Ϫ42 Ϫ48 Ϫ53 Ϫ59 Ϫ64 Ϫ69 Ϫ25 Ϫ27 Ϫ28 Ϫ35 Ϫ41 Ϫ48 Ϫ54 Ϫ60 Ϫ65 Ϫ71 Ϫ76 Ϫ30 Ϫ32 Ϫ33 Ϫ40 Ϫ47 Ϫ53 Ϫ60 Ϫ66 Ϫ72 Ϫ77 Ϫ83 Ϫ35 Ϫ37 Ϫ38 Ϫ46 Ϫ53 Ϫ59 Ϫ66 Ϫ72 Ϫ78 Ϫ84 Ϫ89 Ϫ40 Ϫ42 Ϫ44 Ϫ51 Ϫ58 Ϫ65 Ϫ72 Ϫ78 Ϫ84 Ϫ90 Ϫ96 Ϫ45 Ϫ47 Ϫ49 Ϫ56 Ϫ64 Ϫ71 Ϫ78 Ϫ84 Ϫ91 Ϫ97 Ϫ103 Approximate warming effect (ЊF) of extra radiation (100 W mϪ2, temp ϭ 32ЊF)* 111098765554

* Extra radiation received from sources outside the body.

TABLE 5. NWS (1992b) wind chill (ЊF, mph) per formula in NWS Ops Man. C-42. U.S. units Wind speed (mph) Dry bulb (ЊF) 4 5 10 15 20 25 30 35 40 45 35 35 32 22 16 11 8 6 4 3 2 30 30 27 16 9 4 1 Ϫ2 Ϫ4 Ϫ5 Ϫ6 25 25 22 10 2 Ϫ3 Ϫ7 Ϫ10 Ϫ12 Ϫ13 Ϫ14 20 20 16 4 Ϫ4 Ϫ10 Ϫ14 Ϫ17 Ϫ19 Ϫ21 Ϫ22 15 15 11 Ϫ2 Ϫ11 Ϫ17 Ϫ22 Ϫ25 Ϫ27 Ϫ29 Ϫ30 10 10 6 Ϫ9 Ϫ18 Ϫ24 Ϫ29 Ϫ32 Ϫ35 Ϫ37 Ϫ38 5 5 1 Ϫ15 Ϫ25 Ϫ31 Ϫ36 Ϫ40 Ϫ43 Ϫ45 Ϫ46 0 0 Ϫ5 Ϫ21 Ϫ31 Ϫ39 Ϫ44 Ϫ48 Ϫ51 Ϫ53 Ϫ54 Ϫ5 Ϫ5 Ϫ10 Ϫ27 Ϫ38 Ϫ46 Ϫ51 Ϫ55 Ϫ58 Ϫ60 Ϫ62 Ϫ10 Ϫ10 Ϫ15 Ϫ33 Ϫ45 Ϫ53 Ϫ59 Ϫ63 Ϫ66 Ϫ68 Ϫ70 Ϫ15 Ϫ15 Ϫ20 Ϫ39 Ϫ51 Ϫ60 Ϫ66 Ϫ71 Ϫ74 Ϫ76 Ϫ78 Ϫ20 Ϫ20 Ϫ26 Ϫ46 Ϫ58 Ϫ67 Ϫ73 Ϫ78 Ϫ82 Ϫ84 Ϫ86 Ϫ25 Ϫ25 Ϫ31 Ϫ52 Ϫ65 Ϫ74 Ϫ81 Ϫ86 Ϫ89 Ϫ92 Ϫ94 Ϫ30 Ϫ30 Ϫ36 Ϫ58 Ϫ72 Ϫ81 Ϫ88 Ϫ93 Ϫ97 Ϫ100 Ϫ101 Ϫ35 Ϫ35 Ϫ41 Ϫ64 Ϫ78 Ϫ88 Ϫ96 Ϫ101 Ϫ105 Ϫ108 Ϫ109 Ϫ40 Ϫ40 Ϫ47 Ϫ70 Ϫ85 Ϫ95 Ϫ103 Ϫ109 Ϫ113 Ϫ116 Ϫ117 Ϫ45 Ϫ45 Ϫ52 Ϫ76 Ϫ92 Ϫ103 Ϫ110 Ϫ116 Ϫ120 Ϫ123 Ϫ125

Unauthenticated | Downloaded 10/03/21 04:00 PM UTC DECEMBER 1998 NOTES AND CORRESPONDENCE 1193

TABLE 6. Difference (ЊF): NWS (1992b) from Table 5 minus Steadman wind chill from Table 4. U.S. units Wind speed (mph) Dry bulb (ЊF) 4 5 10 15 20 25 30 35 40 45 35 1 Ϫ1 Ϫ7 Ϫ9 Ϫ10 Ϫ10 Ϫ9 Ϫ7 Ϫ6 Ϫ4 30 1 Ϫ1 Ϫ8 Ϫ10 Ϫ11 Ϫ11 Ϫ10 Ϫ9 Ϫ7 Ϫ5 25 1 Ϫ1 Ϫ8 Ϫ12 Ϫ13 Ϫ12 Ϫ12 Ϫ10 Ϫ8 Ϫ6 20 1 Ϫ1 Ϫ9 Ϫ13 Ϫ14 Ϫ14 Ϫ13 Ϫ11 Ϫ9 Ϫ7 15 1 Ϫ1 Ϫ10 Ϫ14 Ϫ15 Ϫ15 Ϫ14 Ϫ13 Ϫ11 Ϫ8 10 1 Ϫ2 Ϫ11 Ϫ15 Ϫ17 Ϫ17 Ϫ16 Ϫ14 Ϫ12 Ϫ10 5 2 Ϫ2 Ϫ12 Ϫ16 Ϫ18 Ϫ18 Ϫ17 Ϫ16 Ϫ13 Ϫ11 0 2 Ϫ2 Ϫ13 Ϫ17 Ϫ19 Ϫ20 Ϫ19 Ϫ17 Ϫ15 Ϫ12 Ϫ5 2 Ϫ2 Ϫ13 Ϫ19 Ϫ21 Ϫ21 Ϫ20 Ϫ19 Ϫ16 Ϫ13 Ϫ10 2 Ϫ2 Ϫ14 Ϫ20 Ϫ22 Ϫ23 Ϫ22 Ϫ20 Ϫ17 Ϫ14 Ϫ15 2 Ϫ2 Ϫ15 Ϫ21 Ϫ24 Ϫ24 Ϫ23 Ϫ21 Ϫ19 Ϫ15 Ϫ20 2 Ϫ2 Ϫ16 Ϫ22 Ϫ25 Ϫ26 Ϫ25 Ϫ23 Ϫ20 Ϫ17 Ϫ25 2 Ϫ3 Ϫ17 Ϫ23 Ϫ26 Ϫ27 Ϫ26 Ϫ24 Ϫ21 Ϫ18 Ϫ30 2 Ϫ3 Ϫ18 Ϫ25 Ϫ28 Ϫ29 Ϫ28 Ϫ26 Ϫ23 Ϫ19 Ϫ35 2 Ϫ3 Ϫ18 Ϫ26 Ϫ29 Ϫ30 Ϫ29 Ϫ27 Ϫ24 Ϫ20 Ϫ40 2 Ϫ3 Ϫ19 Ϫ27 Ϫ30 Ϫ31 Ϫ31 Ϫ28 Ϫ25 Ϫ21 Ϫ45 2 Ϫ3 Ϫ20 Ϫ28 Ϫ32 Ϫ33 Ϫ32 Ϫ30 Ϫ27 Ϫ22

REFERENCES 5, 20 pp. [Available from NCDC, 151 Patton Ave., Asheville, NC 28801-5001; or [email protected].] Court, A., 1992: Comfort temperatures. McGraw-Hill Encyclopedia Osczevski, R. J., 1995: Comments on ``Wind chill errors'': Part II. of Science and Technology, Vol. 4, McGraw-Hill, 217±218. Bull. Amer. Meteor. Soc., 76, 1630±1631. Dixon, J. C., and M. J. Prior, 1987: Wind-chill indicesÐA review. Schwerdt, R. W., 1995: Comments on ``Wind chill errors'': Part III. Meteor. Mag., 116, 1±16. Bull. Amer. Meteor. Soc., 76, 1631±1636. Driscoll, D. M., 1985: Human health. Handbook of Applied Mete- Siple, P. A., and C. F. Passel, 1945: Measurements of dry atmospheric orology, D. D. Houghton, Ed., John Wiley and Sons, 778±814. cooling in sub-freezing temperatures. Reports on scienti®c re- , 1992: Thermal comfort indexes: Current uses and abuses. Natl. sults of the United States Antarctic Service Expedition, 1939± Wea. Dig., 17, 33±38. 1941. Proc. Amer. Philos. Soc., 89, 177±199. , 1994: Comments on ``Wind chill errors.'' Bull. Amer. Meteor. Steadman, R. G., 1971: Indices of wind chill of clothed persons. J. Soc., 75, 445. Appl. Meteor., 10, 674±683. Horstmyer, S., 1994: Comments on ``Wind chill errors'' by Edwin , 1979a: The assessment of sultriness. Part I: A temperature- Kessler. Bull. Amer. Meteor. Soc., 75, 445±447. humidity index based on human physiology and clothing science. Kessler, E., 1993: Wind chill errors. Bull. Amer. Meteor. Soc., 74, J. Climate Appl. Meteor., 18, 861±873. 1743±1744. , 1994: Reply to Driscoll and Horstmeyer. Bull. Amer. Meteor. , 1979b: The assessment of sultriness. Part II: Effects of wind, Soc., 75, 447. extra radiation and barometric pressure on apparent temperature. , 1995: Reply. Bull. Amer. Meteor. Soc., 76, 1637±1638. J. Climate Appl. Meteor., 18, 874±885. Kuhl, J. K., 1992: Metric wind chill chart and wind speed chart. NWS , 1984: A universal scale of apparent temperature. J. Climate Tech. Procedures Bull. 165, 2 pp. [Available from NCDC, 151 Appl. Meteor., 23, 1674±1687. Patton Ave., Asheville, NC 28801-5001; or [email protected].] , 1994: Norms of apparent temperature in Australia. Aust. Me- NWS, 1992a: Non- weather hazards (C-44). WSOM Is- teor. Mag., 43, 1±16. suance 92-6, 16 pp. [Available from NCDC, 151 Patton Ave., , 1995: Comments on ``Wind chill errors'': Part I. Bull. Amer. Asheville, NC 28801-5001; or [email protected]]. Meteor. Soc., 76, 1628±1630. , 1992b: Winter weather warnings (C-42). WSOM Issuance 92- WMO, 1997: The German Klima±Michel model. WMO Bull., 46, 31.

Unauthenticated | Downloaded 10/03/21 04:00 PM UTC