Mc. Acock et Ya. Pachepsky, Estimating missing weather data for agricultural simulations using group method of data handling, J APPL MET, 39(7), 2000, pp. 1176-1184
Contacting weather stations via modems to obtain weather data for crop simu
lations has become a common practice. Users sometimes encounter gaps in the
se data, and techniques are needed to estimate weather variables for days w
hen the data are absent. The authors hypothesized that such estimations can
be made using data from before and after the day with no data. Dependencie
s of weather variables of a particular day on weather variables from severa
l days before and after could be very complex. To find and to express these
dependencies, group method of data handling (GMDH), which is a tool for mo
deling complex "input-output" relationships by building hierarchical polyno
mial regression networks, was used. Data on daily solar radiation, maximum
and minimum temperatures. and wind runs collected daily in Stoneville, Miss
issippi, during May-September of 1982-92 were used. Fourteen-hundred sequen
tial 7-day datasets from the database were extracted. For each dataset, the
authors assumed that weather variables on the fourth day were unknown and
had to be found from the weather variables of days 1, 2, 3, 5, 6, and 7. Se
venty-five percent of these data were used to iind the hierarchical polynom
ial regression, and 25% were used to evaluate it. Correlation coefficients
between calculated and actual parameters were similar for training and eval
uation datasets. Coefficients of determination (R-2) were about 0.88 for mi
nimum temperature. 0.80 for maximum temperature, and 0.80 for wind run. Acc
uracy of the solar radiation and precipitation estimates was lower, and R-2
was about 0.2-0.3 but improved to 0.5-0.6 for the training dataset and 0.3
for the validation dataset for both variables when an additional indicator
variable that shows the presence or absence of rain was included. The next
day after the day with missing data gave the most essential information. I
ncreasing the number of missing days resulted in gradual deterioration of t
he accuracy for ail variables but wind run. GMDH fan be a useful tool for f
illing gaps in weather data from weather stations installed in the field.