Bu çalışma, ayl ı k ak ı mları n modellenmesi amac ı yla yap ı lmışt ı r. Ayl ı k ak ı mları n modellenmesinde ARMA p,q modelleri ile Thomas-Fiering modeli kullan ı lm ıştı r. ARMA p,q modellerinin seçiminde, ARMA p,q modellerinden hesaplanan kal ı ntı ları n bağı ms ı z olup olmad ı kları göz önüne al ı nmıştı r. Ayl ı k ak ı mlar için, ARMA 1,0 , ARMA 2,0 , ARMA 1,1 , ARMA 1,2 ve ARMA 3,0 modelleri uygun olmu ştur. Ancak ayl ı k ak ı mlara ARMA 2,0 modeli daha fazla uyum göstermiştir. ARMA 2,0 modeli ve Thomas-Frieng modelinden tahmin edilen ayl ı k ak ı mlar ile gözlenen ak ı mlar, korelasyon katsay ı ları na göre karşı laştı rı lmıştı r. ARMA 2,0 modeli için bu katsay ı lar daha yüksek olmu ştur. Bu nedenle ayl ı k ak ı mları n tahmininde ARMA 2,0 modeli uygun bulunmu ştur.
This study was made to model monthly flows. ARMA p,q models and Tomas-Fiering model were used in modeling monthly flows. In selecting ARMA p,q models, it was taken into account whether the residuals calculated from ARMA p,q models were independent. For monthly flows, ARMA 1,0 , ARMA 2,0 , ARMA 1,1 , ARMA 1,2 ve ARMA 3,0 models were appropriate. But, ARMA 2,0 model was fitted better to monthly flows. The monthly flows forecasted from ARMA 2,0 model and Thomas-Fiering model were compared to the observed flows considering their correlation coefficients. The correlation coefficients for ARMA 2,0 model were very high. Therefore, ARMA 2,0 model in forecasting monthly flows was preferred.
Primary Language | Turkish |
---|---|
Journal Section | Research Article |
Authors | |
Publication Date | May 1, 2003 |
Submission Date | May 1, 2003 |
Published in Issue | Year 2003 Volume: 09 Issue: 02 |
Journal of Agricultural Sciences is published open access journal. All articles are published under the terms of the Creative Commons Attribution License (CC BY).