Continuous real-time health monitoring in animals is essential for ensuring animal welfare. In ruminants like cows, rumen health is closely intertwined with overall animal health. Therefore, in-situ monitoring of rumen health is critical. However, this demands in-body to out-of-body communication of sensor data. In this paper, we devise a method of channel modeling for a cow using experiments and FEM based simulations at 400 MHz. This technique can be further employed across all frequencies to characterize the communication channel for the development of a channel architecture that efficiently exploits its properties.