Physically-based Distributed Hydrologic Modeling - PowerPoint PPT Presentation

PPT – Physically-based Distributed Hydrologic Modeling PowerPoint presentation | free to download - id: 69d43f-NDUyN

The Adobe Flash plugin is needed to view this content

Get the plugin now

View by Category
Title:

Physically-based Distributed Hydrologic Modeling

Description:

Physically-based Distributed Hydrologic Modeling – PowerPoint PPT presentation

Number of Views:2
Avg rating:3.0/5.0
Slides: 17
Provided by: seasUclaE
Category:
Transcript and Presenter's Notes

Title: Physically-based Distributed Hydrologic Modeling

1
Physically-based Distributed Hydrologic Modeling
2
Goal of Phys.-based Distrib. Hydrologic Modeling
• To date we have learned about
• Key forcings at land surface (precipitation/net
• Physical processes at surface/subsurface
(infiltration, soil moisture redistribution,
evapotranspiration, groundwater flow, runoff,
etc.)
• Goal Develop physically-based model of
hydrologic response across a watershed by tying
together various processes across landscape.
• In this context Distributed refers to variables
being spatially-distributed in space.
• So we aim to explicitly model how the hydrologic
states/fluxes evolve in space and time throughout
the watershed.
• Note Because of complexity/nonlinearity of
processes this modeling is necessarily done
numerically (i.e. by building appropriate
computer models coupling together hydrologic
processes)

3
Representation of Dist. Hydrologic Units in
Space
Numerical simulations of catchment hydrologic
processes require a method for representing a
basin. Methods can be categorized as lumped
versus distributed modeling where the physical
processes are solved for each discrete unit.
Basin-Averaged Models (e.g. HEC-HMS)
Raster (Grid) Models (e.g. MIKE SHE)
Triangular Irregular Network Models
(e.g. tRIBS)
Will focus on this model as an example
4
tRIBS Distributed Model
TIN-based Real-time Integrated Basin Simulator
(tRIBS) is a fully-distributed model of coupled
hydrologic processes (Ivanov et al, Vivoni et al.)
• Model Processes
• Coupled vadose and saturated zones with dynamic
water table.
• Moisture infiltration waves.
• Soil moisture redistribution.
• Topography-driven lateral fluxes in vadose and
groundwater.
• Evaporation and Transpiration.
• Hydrologic and hydraulic routing.

Key point You now know about all of these
processes a distributed model simply ties them
all together.
5
Process Representation Surface Processes
Land-Atmosphere Interactions
• Coupled Energy and Hydrology Processes on Complex
Terrain
• Radiation Incoming short-wave and long-wave,
of terrain).
• Vegetation Canopy interception, drainage,
throughfall and transpiration using vegetation
functional type.
• Energy Balance Net radiation, ground heat,
sensible heat and latent heat fluxes.
• Evapotranspiration Soil-moisture controlled
bare soil evaporation and canopy transpiration in
root zone.
• Unsaturated Zone Dynamics Soil moisture
balance, infiltration, redistribution

Vegetation
3D Complex Topography
Soil
Surface Energy Balance
Aquifer
6
Process Representation Subsurface Processes
Uses a simplified 2D unconfined aquifer model
which allows moisture recharge in shallow aquifer
to be redistributed.
• Shallow Groundwater
• Space/time variable groundwater table position.
• Single and multiple direction GW flow to
downstream neighbors.
• Coupled to unsaturated zone to enable moisture
mass balance (recharge).
• Bounded by a uniform or spatially-variable
bedrock surface (impermeable bottom boundary).

Variable, dynamic water table field (plan view)
7
Process Representation Unsat.-Sat. Dynamics
Runoff is generated via multiple mechanisms
depending on the interactions of infiltration
fronts and the water table.
• Runoff Generation
• Interaction of rainfall, infiltration capacity,
actual infiltration and lateral flows lead to
various runoff types.
• Various runoff types occur at the same time in
different basin parts.
• Various runoff types can occur in single element
as a function of state.
• Infiltration-excess (Hortonian) Runoff.
• Saturation-excess (Dunne) Runoff.
• Perched Subsurface Runoff.
• Groundwater Runoff.

Example Model output for saturation-excess
runoff occurrence
8
Atmospheric Forcing
Primary reason for using distributed models is to
take advantage of new distributed atmospheric
etc).
9
tRIBS Model Output
• tRIBS provides output at the scale of each
individual node in the basin, for channel nodes
along the network, and as maps of distributed
variables (at a point in time or integrated over
time).
• Time Series of Node Behavior Unsaturated and
Saturated Node Dynamics, Hydrologic and Energy
Fluxes and State Variables.
• Basin Outlet and Interior Channel Nodes Runoff
Depth, Discharge, Stream Velocity, Partitioned
Hydrographs.
• Dynamic Distributed Maps Groundwater dynamics,
Surface Runoff Generation Mechanisms, Soil
Moisture, Evapotranspiration, Rainfall,
Interception, Unsaturated Zone Dynamics, Energy
• Integrated Distributed Maps Percent Runoff
Mechanisms, Saturation Occurrence, Evaporation
Fraction, Soil Moisture.
• Time Series of Basin Averaged Properties
Rainfall, Saturated Area, Evapotranspiration,
Soil Moisture.

lumped model!
10
Illustrative Example Peacheater Creek
Two-year precipitation record
11
Parameter Definitions for Basin
(silt loam)
(mixed forest)
(everg. forest)
(decid. forest)
(silty clay)
(crops)
(clay/ urban)
(urban)
Note Spatially varying inputs in soil/vegetation
-- impacts spatial variability in hydrologic
response
12
Streamflow Response (Storm at Hour 11800)
13
Groundwater Before/after
14
Soil Moisture Before/after
15
Surface Energy Balance
16
Summary
• Distributed hydrologic modeling provides an
integrated framework for taking into account
hydrologic processes occurring within the basin
(surface energy balance, flow partitioning, etc.)
• Allows for not only simulating design flows/flood
forecasts (i.e. as done using UH-method), but for
things like assessing spatial response to inputs,
hydrologic impacts resulting from urbanization of
watersheds, assessing climatology of hydrologic
states, etc.
• Takes advantage of many new distributed
forcing/parameter databases obtained via remote
sensing (lumped models do not take advantage of
spatially distributed inputs)
• Is computationally demanding (e.g. compared to
UH) and therefore whether it should be used is
largely application dependent