Yield forecasting for olive tree by meteorological factors and pollen emission

• Main Authors: Aferdita LASKA MERKOCI, Albana HASIMI, Mirela DVORANI
• Format: Article
• Language: English
• Published: Geo-SEE Institute 2019-06-01
• Series: Micro, Macro & Mezzo Geoinformation
• Subjects: equations of regression; forecasting; meteorological factor; olive; yield product
• Online Access: http://mmm-gi.geo-see.org/wp-content/uploads/MMM-GI_12/Laska-Merkoci_A-Hasimi_A-Dvorani_M.pdf
• ISSN: 1857-9000; 1857-9019

The paper aims to forecast the olive product based on the application of a statistical model by use of meteorological factors and pollen emission. Nowadays there are a number of models and approaches related to the yield forecasting. All of them have their advantages and disadvantages and moreover different behaviours for climate conditions of Albania. Thus, after a preliminary evaluation the best fitted model was chosen and its result were analysed. The model was based on the multiple equations of regression, which took into consideration some climate factors. These factors are rainfall in May followed by rainfall in June. Minimum temperatures during spring and summer were also an important consideration due to the influence of night temperature on energy collected for fruit development. The use of pollen emission and monthly meteorological data from 1985-2004 as predictive variables has enabled the production of a forecast up to 8 month prior to the end of harvesting. The forecasting of yield production in this study has been made in November, which reflects the EPP and the meteorological factors like minimum temperature, maximum temperature, rainfall from May to October etc. In addition, as the model requires, the most significant periods for this plant were chosen and evaluated for the Vlora region of Albania with the highest productivity in the country. Results were compared with real olive crop data and estimates from the equation resulted to have a correlation coefficient about 0.77 and SE=3.0.