Estimation of Crop Coefficient of Corn (Kccorn) under Climate Change Scenarios Using Data Mining Technique

Abstract

The main objectives of this study are to determine the crop coefficient of corn (Kccorn) using data mining technique under climate change scenarios, and to develop the guidelines for future water management based on climate change scenarios. Variables including date, maximum temperature, minimum temperature, precipitation, humidity, wind speed, and solar radiation from seven meteorological stations during 1991 to 2000 were used. Cross-Industry Standard Process for Data Mining (CRISP-DM) was applied for data collection and analyses. The procedures compose of investigation of input data, model set up using Artificial Neural Networks (ANNs), model evaluation, and finally estimation of the Kccorn. Three climate change scenarios of carbon dioxide (CO2) concentration level: 360 ppm, 540 ppm, and 720 ppm were set. The results indicated that the best number of node of input layer - hidden layer - output layer was 7-13-1. The correlation coefficient of model was 0.99. The predicted Kccorn revealed that evapotranspiration (ETcorn) pattern will be changed significantly upon CO2 concentration level. From the model predictions, ETcorn will be decreased 3.34% when CO2 increased from 360 ppm to 540 ppm. For the double CO2 concentration from 360 ppm to 720 ppm, ETcorn will be increased 16.13%. The future water management guidelines to cope with the climate change are suggested.

Keywords

• Crop Coefficient
• Crop Evapotranspiration
• Climate Change
• Artificial Neural Network
• Data Mining