Bayesian Network Modeling Approach Assessing Floodplain Resilience of the Gila River, New Mexico, USA
Abstract
A Bayesian network (BN) approach is implemented to understand riparian vegetation resilience for the Upper Gila Basin in southern New Mexico, USA. We implement both two-dimensional hydrodynamic and Bayesian network modeling to incorporate spatial variability in the system. Our coupled modeling framework presents vegetation recruitment potential for regional representative plant types under several hydrologic scenarios and management actions. In our BN model, we incorporate key ecological drivers that address timing, floodplain inundation, river recession rate, and groundwater conditions as well as constraints for the plant types based on expert knowledge and literature. Our approach allows us to utilize small and incomplete data, incorporate expert knowledge, and explicitly account for uncertainty in the system. We were able to evaluate riparian vegetation consequences of water management scenarios at a fine spatial scale, which helped identify distinctly impacted channel locations, and can provide local decision-makers insights to guide risk management, watershed restoration activities, and future development project that promote watershed health and floodplain resilience.
- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2019
- Bibcode:
- 2019AGUFM.H53L1938J
- Keywords:
-
- 1813 Eco-hydrology;
- HYDROLOGY;
- 1820 Floodplain dynamics;
- HYDROLOGY;
- 1890 Wetlands;
- HYDROLOGY;
- 4327 Resilience;
- NATURAL HAZARDS