A Comparison of Deterministic and Stochastic Models for Predicting Air and Litter Properties in a Broiler Building

Abstract

This paper presents a comparison of deterministic interpolation methods, including inverse distance weighting (IDW; power 1, 2, and 3) and radial basis function (RBF; basis function: thin-plate spline, natural cubic spline, multilogarithmic function, multiquadric function, inverse multiquadric function, and stochastic interpolation methods, including ordinary kriging (OK; semivariogram model: spherical, exponential, and Gaussian) to determine the best method to estimate indoor air environmental conditions and the physical properties of litter in a broiler building. Measurements were taken at the beginning (Day 7), middle (Day 21), and end (Day 40) of the summer and winter flocks in the building, in which 80 sampling positions composed an imaginary grid. Results indicate that the OK perform overall better than the deterministic methods. Geostatistical maps created using the optimal interpolation method (OK-Gaussian) were employed to infer the regions of stress and comfort situations for both seasons. It was concluded that indicate adequate cooling was not provided during the end of the summer flock. In the winter flock, higher litter moisture levels were observed near the building's ends and along the sidewalls. The results of rigorous sampling of the winter and summer flocks, visually reported using geostatistical maps, can assist the poultry industry and researchers in improving planning and management strategies.