Auto-calibrate Hargreaves model

Description

Function estimates Hargreaves model coefficients 'A' and 'B' based on autocalibration procedure

Usage

1
2
hauto(lat, lon, days,extraT = NULL, Tmax, Tmin, tal,
Ha_guess = 0.16, Hb_guess = 0.1, epsilon=0.5, perce = NA) 

Arguments

lat

Latitude in decimal degrees.

lon

Longitude in decimal degrees.

days

Vector of class 'Date' of length n.

extraT

Optional. Vector of length n of extraterrestrial solar radiation [MJm-2]. If 'NULL' then it is calculated by the function. Providing extraterrestrial solar radiation speeds up the computation

Tmax

Vector of length n containing daily maximum temperature [C].

Tmin

Vector of length n containing daily minumum temperature [C].

tal

Clear sky transmissivity.

Ha_guess

Assumption of Hargreaves Ha coefficient. Default set to 0.16.

Hb_guess

Assumption of Hargreaves Hb coefficient. Default set to 0.1.

epsilon

A value of which potential radiation is decreased. See "details".

perce

Percent of clear days. Default set to 1.

Details

The auto-calibration method bases on the assumption that on the clear-sky days model should not overpredict potential values. To define those clear-sky days, we estimate daily solar radiation using Hargreaves model with default values of A = 0.16, B = 0.1 and tal = 0.72 and we select those days for which estimated daily solar radiation is the closest to the potential values (extra-terrestrial*tal). The number of clear-sky days is estimated based on the mean Cloud Fraction Cover map. Next, based on selected clear-sky days, we perform a non-linear least squares regression to derive A and B coefficients treating potential values decreased by 'epsilon' as a reference solar radiation values. The analysis of auto-calibration results showed clear correlation between optimal 'epsilon' and distance to the coast. We proposed simplified method in which 'epsilon' is equal to 0.1 MJm-2 or to 0.5 MJm-2 when distance to the coast is smaller or bigger than 15 km respectively.

Value

Vector of length 3 containing:

Ha

Hargreaves 'A' coefficient

Hb

Hargreaves 'B' coefficient

Hr2

Coefficient of determination of performed linear regression

Author(s)

Jedrzej S. Bojanowski

References

Hargreaves, G.H., Samani, Z.A. 1892. Estimating potential evapotranspiration. J. Irrig. Darin. Eng., ASCE 108 (3), 225-230. Bojanowski, J.S., Donatelli, M., Skidmore, A.K., Vrieling, A., 2013. An auto-calibration procedure for empirical solar radiation models Environmental Modelling and Software 49, 118-128.

See Also

'hacal'

Examples

1
2
3
4
5
6
7
data(Metdata)
Tmax <- Metdata$meteo$TEMP_MAX
Tmin <- Metdata$meteo$TEMP_MIN
days <- Metdata$meteo$DAY
lat <- Metdata$LATITUDE
lon <- Metdata$LONGITUDE
hauto(lat,lon,days,extraT=NULL,Tmax,Tmin,tal=0.76)