A Survey of Major Watershed (Rainfall- Runoff) Models z Knowing what is out there in the Workplace z...

Post on 27-Mar-2015

217 views 1 download

Tags:

Transcript of A Survey of Major Watershed (Rainfall- Runoff) Models z Knowing what is out there in the Workplace z...

A Survey of Major Watershed (Rainfall-Runoff) Models

Knowing what is out there in the Workplace

Capabilities of each model Requirements of each model What is the best model? Examples

Ungaged area needing WRE management (expansion)

Balancing $$, needs and requirements (Consulting)

Flood-frequency analyses, forecasting, extrapolating for future scenarios

Consider this Scenario:

1. A major expansion of urbanization is planned in forested land – malls, apts, parks, industry

2. How will the planned urbanization change the response of the land to rainfall? Increase chances of floods? Flashier? Bring down GWT?

What kind of Water Resources Management Strategy/Measures would be needed to ensure long-term water needs?

Watershed Models are used to answer such questions

GOALS

Be aware of the Major Models out there and their basic workings

Understand Data needs and requirements Merits/Demerits of Models (situational)Ponder over: What is the Best Model?

(open question) Be able to make an informed decision on the

choice of a Model (for Consulting, Industry, Public Works)

The Hydrologic Cycle

All watershed Models (or Rainfall-Runoff) attempt to model the Hydrologic Cycle at the watershed scale.

Watershed? – Natural System with all major components of cycleHydrologic Cycle? (Water Cycle) – Water DistributionModel? – Abstraction of reality using Mathematics

The Major Processes of a Model

Infiltration

Unsaturated Zone flow

Groundwater (saturated

zone flow)

Overland Flow

River Flow

Evapotranspiration

Rainfall is usually the major input in Hydrologic Models – it is not modeled!

Classification of Models

Based on mathematical representation of processes

Conceptual Models Physically-based Models

Based spatial nature of input/output of modelsLumped Rainfall, Runoff, ET, GW flowDistributed Rainfall, Runoff, ET, GW

flow

Conceptual Vs Physically Based

Conceptual Models – Idealization of processes as stores, buckets, parameterizations – simplified equations representing mass, momentum, energy.

Physically-based Models – “rigorous numerical solution of partial differential equations governing flow through porous media, overland and channel flows. “

Lumped Vs Distributed Models

Distributed - The spatial and temporal variation of meteorological, hydrological, geological and hydrogeological data across the model area is described in gridded form for the input as well as the output from the model

Lumped - The spatial and temporal variation of meteorological, hydrological, geological and hydrogeological data across the model area is described as one aggregated value for input and output

!All models are lumped at the finest scale!

Data Requirements for Models

Rainfall Data (Major Input) Soils Data (Infiltration,

Runoff) DEM – channel network

(River routing) Vegetation Data (For ET) GWT Data (Saturated zone

flow) Historical Rainfall-

Streamflow Data (Calibration)

Evaporation Data (ET)

More Physically-based means more data requirement

More Conceptual requires less data!

Data acquisition is an investment that needs to be commensurate with the model

Sources of Data

Rainfall – Gage, Radar (NWS), Satellite

Soils – STATSGO, USDA

DEM – USGS, Topo maps, Satellites

Vegetation – Satellites, USDA

Streamflow – USGS

ReliabilityReady availabilityCost $$$Resolution (space,

time)

Is all data needed?

Physically-based – Insatiable needs – but potential high returns

Conceptual – modest needs with modest returns

Calibration –needs historical data. What is Calibration?

What to do in ungaged regions where there is ‘No Data’? Transfer experience. PUB

What affects $$ operational costs of a Model?

The multiple needs of the TVA river reservoir system

Level of data needs – Data is expensive Level of physical complexity (Physically-

based means qualified personnel; Rational Method – no-brainer

Scale of the problem being addressed – larger/heterogenous – more $$$

Some Models Out there

HEC-1 PRMS HSPF (Fortran) MIKE 11 Make your own survey if you wish

(Consider, DSS – Riverware, BASINS)

For a comprehensive guide refer to: VP Singh – Computer Models for Watersheds

HEC-1

Developed by Hydrologic Engineering Center (HEC) of US Army Corps of Engineers

Flood Hydrograph Package – single storm even simulation. Lumped model

Loss Function approach – SCS, Green Ampt, Holtan Data Needs – sub-basin delineation, rainfall, runoff, routing

parameters Based on Unit Hydrograph or kinematic for runoff routing Other capabilities – Snowfall/melt, dam safety, pumping,

diversions Customer Support? – Try HEC, San Diego, CA.

PRMS – Precipitation-Runoff Modeling System

Developed by USGS – more of an educational tool to build your own models

Lumped/sub-basin –continuous simulation – see Handout

Data needs – (depends) precip, streamflow, DEM, minmax air temp, radiation, vegetation..(see handout)

Code in F77Capabilities - manyCustomer Support? – Try USGS

HSPF Hydrological Simulation Program - Fortran

USGS – multi-use modelCan do water quality.Mainly for land-use change, reservoir

operations, flow diversions etc.Data needs – A lot – (see Hand-out)Software – In Fortran 77Capabilities – A lot (see hand-out)Customer Support? – Try USGSReal Example – 62,000 sq mi tributary

area in Chesapeake Bay.

MIKE 11

Danish Hydraulic Institute DHIVery sophisticated – physically-based,

high-end modelData needs – a lot (dependent of in-situ

measurementsSoftwares – commercial (and expensive)Capabilities – Many (see hand-out)Customer Support? – Try DHI or VendorsReal Example – River Jamuna, Bangladesh

MIKE SHE

USGS – educational, conceptual/physical

Data needs?SoftwaresCapabilitiesMerits/Demerits (can/can not do)Customer Support?Real Example -

What is the best Model?

This is situational – It All Depends!

Primary objectives

$$ budget

Level of Tolerable Accuracy/precision

Complexity and uniqueness of the problem

Time frame for delivery

Data availability

It’s a highly complex question whose answer is sensitive to a multiplicity of factors

Homework Problem

Given watershed,Given dataGiven $$, deadlineChoose your model (make strong

arguments in favor of your choice over others)