Flume calibration on irrigated systems by Video Image Processing and Bayesian Inference
Rice is a crop that requires a large amount of water throughout its production cycle to ensure a good yield, resulting in higher water consumption compared to other crops. In Uruguay, about 160,000 ha/year are planted, requiring about 1,760 hm3/year of water, with a very high international average productivity of 9,000 kg/ha. Irrigation is generally carried out by surface/flooding, with water passing through dug channels where gates are used to regulate the flow, and in some cases measurement devices are installed. Increasing pressure on water resources makes it necessary to increase knowledge of water use at the farm level. Flumes are an opportunity in this sense; however, they require calibration and adjustment through gauging, which are generally omitted due to their high cost and complexity. In this work, an economic method for the calibration of flumes through image video processing is proposed. The method uses the RIveR software (https://riverdischarge.blogspot.com/) for the video image processing, and the BaRatinAGE software to establish the stage-discharge relationship through Bayesian inference. A Surface Velocity Radar and an Acoustic Doppler Velocity Meter are used as reference sensors. The methodology is tested on a cutthroat flume. The experiment was conducted at a rice farm in northern Uruguay. The results indicate that flumes can be easily calibrated by video image processing and uncertainty can be quantified through Bayesian inference. An advantage of the proposed method is that it uses free software that can be easily applied in small farms.
| Main Authors: | , , , , , |
|---|---|
| Format: | Digital revista |
| Language: | spa |
| Published: |
Coeditada entre Facultad de Agronomía - Udelar y el Instituto Nacional de Investigación Agropecuaria (INIA)
2024
|
| Online Access: | https://agrocienciauruguay.uy/index.php/agrociencia/article/view/1182 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|