where *M *= snowmelt (in. day1)

Output of modeled snow information in

HEC-1 is limited to melt contribution to

"rainfall" excess. Temporal changes in the

(dimensionless)

snowpack depth or snow water equivalent

are not given.

(dimensionless),

Temperature lapse rates are fixed and can-

not be varied with actual weather condi-

tal surface (langley day1),

tions.

α = albedo (dimensionless),

(dimensionless)*,*

SSARR (Streamflow Synthesis and Reservoir

Regulation) was developed by the North Pacific

snow surface temperatures (F),

Division (NPD) Corps of Engineers beginning in

1956 to provide hydrologic simulations on snow-

surface snow temperatures (F),

melt-dominated river systems for planning,

design, and operation of water control works (U.S.

for variation from the generalized

Army Corps of Engineers 1991). SSARR was later

snowmelt equation (dimensionless).

expanded to provide operational river forecasting

Term 1 represents melt due to direct solar radia-

and river management for the Columbia River.

SSARR was developed during a time when elec-

melt, term 3 is condensation melt, and term 4 is

tronic digital computers first made continuous

longwave radiation melt in the forest. The short-

stream flow hydrograph simulation practical. The

wave radiation melt factor (*k*′) depends on aver-

philosophy of the model developers was that limi-

age exposure of an open area in comparison to a

tations in data quantity and quality, and in devel-

horizontal surface and is assumed equal to 1.0 in

opment of fundamental relationships, prevented

HEC-1, implying a horizontal surface. The forest

the development of all-purpose, physically based

canopy coverage (*F*) is fixed at 0.5. Albedo (α) is

models. SSARR, thus, was conceptually based and

reduced from 0.75 to a minimum of 0.4, using the

of sufficiently limited detail to allow operational

inverse square of days since the last snowfall to

application on a daily basis. SSARR was one of 11

account for factors that reduce albedo as the snow

models from eight countries evaluated in a world-

ages, such as increased snow grain size. The con-

wide comparison of snowmelt runoff models

densationconvection coefficient (*k*) is taken as 1.0,

(WMO 1986), and has been used extensively in

and the snow surface temperature is taken as 32F.

operational snowmelt modeling in the Pacific

Northwest. The basic snowmelt equations are the

same as those used in HEC-1, though with fewer

The HEC-1 methods do not allow for impor-

restricted coefficients.

tant snowpack processes such as snow rip-

ening, pore water retention, and flow of

water through the pack.

The degree-day, temperature index method is

The use of wind speeds measured at 50 ft

the same equation as used in HEC-1 (eq 1), except

(15.2 m) is awkward, since many weather

that the degree-day melt coefficient (*C*d) can be

stations measure at 2- or 10-m heights. It is

varied during the model run on as much as a daily

uncertain whether methods developed for

basis.

wind speeds at 50 ft convert well to other

measurement heights.

The HEC-1 snow methods have been used

The SSARR energy balance method equations

more often in planning studies than in fore-

are the same as HEC-1 (eq 2 and 3) with the fol-

casting, and in situations where snowmelt

lowing important additions:

runoff is not a primary contributor.

Fractional forest cover canopy (*F*) is not

Radiation, convection, and condensation

fixed, but can be varied from 0 to 1,

coefficients that vary to represent a range of

watershed conditions in EM-1110-2-1406 are

Albedo (α) can be specified.

fixed to midrange values in HEC-1.

3

Back to contents