Hi all,
I am considering replacing the tipping bucket routine in a model and using the low computation “Integrated Richards Equation” approach described in Yang et al (2009) to improve the timing of water flow and the estimation of soil water content, particularly above field capacity. The IRE method solves the Richard’s Equation only for a soil layer, the layer above and the layer below, and then steps through the layers at a sufficiently low time step. The van Genuchten model could be used for characterizing the water retention curve. I can also keep the tipping bucket approach as an option for the user.
Other possible approaches I've been looking into are: Philip’s quasi-analytical method (and similar derivations based off Philip’s) or the Finite water-content vadose zone flow method (https://en.wikipedia.org/wiki/Finite_water-content_vadose_zone_flow_method) but I think they both require a homogeneous soil profile. We recently included an option for a heterogeneous profile in the model (DNDCv.CAN) we are updating.
Any hydrology modelers care to comment?
Thanks,
Ward
Yang, D., T. Zhang, K. Zhang,D. Greenwood, J. P. Hammond, and P. J. White. 2009. An easily implemented agro-hydrological procedure with dynamic root simulation for water transfer in the crop–soil system: Validation and application. Journal of Hydrology 370: 177–190
Hi Ward,
In meteorology the tipping bucket measuring device has an alternative. The RG-11 Rain Gauge senses water hitting its outside surface using beams of infrared light. It uses the same sensing principle used in millions of automotive rain sensing windshield wiper controls. Consequently, it is remarkably inexpensive, at just $59 for a single unit. The RG-11 is optical– not mechanical, chemical, or conductive. Consequently, it is far more rugged, sensitive and reliable than any other technology. The sensor is extremely sensitive, and virtually immune to false trips. Yet, it is completely unaffected by jostling and motion. There are no exposed conductors to corrode, and no openings for bugs to crawl into. There is no place for leaves or other debris to collect.
Concerning modelling an alternative can be found in the Book Hydrology: principles and Processes by Mark Robinson and Roy Ward published by IWA Publishing. It gives an extensive overview of different methodologies for measuring and modelling at the watershed level.
Another goof refernce is: Hydrological Models for Environmental Management
edited by Mikhail V. Bolgov, Lars Gottschalk, Irina Krasovskaia, Robert J. Moore.
Best regards,
Frank
For a water balance model for cereal cropping in a Mediterranean climate I used both 'tipping bucket' and a quasi richards method. The model used soil horizons as layers and operated on a daily step. The approach was to specify a flow rate at field capacity (FC) and assume a miniscule rate at wilting point (0.001mm/day) to define a semilog flow rate curve. On any day the flow could be calculated by the fraction of available water in the layer. The model was tested extensively. It only failed when the water content of a soil layer was significantly higher than FC and produced huge flow rates - this was rare and happened when daily rainfall was very large. In retrospect I should have specified the flow at (say) double FC. Results were much more realistic at negligible computational cost and the cost of one additional data item - which could be estimated from the FC/mm depth (ie soil texture) which the model needed, anyway. Model is published and available - get in touch if you want a copy.
Cliff and Frank, thank you both kindly for these answers. Cliff, I will get in touch since you have clearly thought about this problem and have experimented with a solution.
Best regards,
Ward
- Badges
- Science topic
- Similar topics
- Clinical Pharmacology
- Dosing