Portal de Periódicos da CAPES

Model Prediksi Hasil Panen Berdasarkan Pengukuran Non-Destruktif Nilai Klorofil Tanaman Padi

2019; Gadjah Mada University; Volume: 39; Issue: 4 Linguagem: Inglês

10.22146/agritech.34893

2527-3825

Fitri Hidayah Nasution, Santosa Santosa, Renny Eka Putri,

Data Mining and Machine Learning Applications

Developing yield prediction model was done to predict the production of rice crops based on chlorophyll value at different age levels. A model was developed by using spatial variability data of chlorophyll value. It was measured with a non-destructive method by using chlorophyll meter CCM-200 plus at different age levels in terms of days after planting 25, 40, 60 and 70 (DAP), and yield of rice. The objective of developing yield prediction model was to describe correlation between chlorophyll value and yield at different ages that had been obtained from 20 observation plots at the rice field. The study area was in Banda Langik, Sungai Bangek village, Lubuk Minturun of Koto Tangah sub district in Padang. Data were collected in two systems; grid sampling point and crop cutting test (CCT). Measuring of chlorophyll contained in leaf or number of SPAD (soil plant analysis development) was done by using chlorophyll meter CCM-200 plus. The research showed that chlorophyll value in rice crop correlated with yield. It was proved by correlation index obtained in each stage of age; 25 DAP (r = 0.945), 40 DAP (r = 0.887), 60 DAP (r= 0.835) and 70 DAP (r= 0.897). Rice yield could be predicted through following model: Y = -0.431513 + 0.045144 X1 + 0.03645 X2 + 0.01017 X3 + 0.020551 X4, where Y was the rice yield (kg/m2) and X was chlorophyll value at different age levels (X1=25 DAP), (X2=40 DAP), (X3=60 DAP) and (X4= 70 DAP). The model was produced through a multiple linear regression test based on chlorophyll value data and rice productivity during 1 period of harvest session.