|Although frost can cause considerable crop damage and mitigation practices against forecasted frost exist, frost forecasting technologies have not changed for many years. The paper reports a new method to forecast the monthly number of frost days (FD) for several meteorological stations at Community of Madrid (Spain) based on successive application of two models. The first one is a stochastic model, autoregressive integrated moving average (ARIMA), that forecasts monthly minimum absolute temperature (tmin) and monthly average of minimum temperature (tminav) following Box-Jenkins methodology. The second model relates these monthly temperatures to minimum daily temperature distribution during one month.
|EGU 2009 - 6th General Assembly