TRANSITION FROM LUMPED TO DISTRIBUTED SYSTEMS

1
TRANSITION FROM LUMPED TO DISTRIBUTED SYSTEMS

• Victor Koren, Michael Smith,
• Seann Reed, Ziya Zhang
• NOAA/NWS/OHD/HL, Silver Spring, MD

2
Distributed and Lumped Modeling Dynamics
3
History Lessons

• There is a similarity in dynamics of lumped and
  distributed model developments
• There is a large delay between model development
  and application
• There is no unique best model. Selection for
  application is rather arbitrary process that
  depends on an expertise of the user and practical
  requirements
• Most successful models in an operational use are
  models which have well developed parameterization
  tools

4
Distinguishing Features of Lumped and Distributed
Models

• Physics
• Does point rainfall-runoff model represent well
  field processes
• Can hillslope/channel routing be represented well
  on practically reasonable space/time scales
• Does statistical approach solve a basin
  heterogeneity problem

5
Distinguishing Features of Lumped and Distributed
Models (Continued)

• Physics
• Does statistical approach solve a basin
  heterogeneity problem

Surface runoff simulated with and without use of
rainfall distribution function at different
scales
6
Distinguishing Features of Lumped and Distributed
Models (Continued)

• Space/Time Variability
• Does accounting for the space/time variability of
  input data and parameters guarantee better
  results
• Does scale effect significantly on the model
  structure
• Is a lumped model a reasonable candidate in a
  distributed system

Effect of noisy rainfall data on the peak volume
at different simulation scales.
7
Distinguishing Features of Lumped and Distributed
Models (Continued)

• Parameterization/Calibration
• Can distributed model parameters be measured on
  the grid scale
• Are distributed model parameters identifiable
  enough from hydrograph analyses
• How much does scale effect on model parameters

Change an effective parameter value at
different scales as a function of
rainfall variability
8
HL-Research Modeling System (HL-RMS)

• Modeling framework for testing lumped,
  semi-distributed, and fully distributed
  hydrologic modeling approaches

9
HL-RMS Structure

• Uses channel connectivity matrix defined on the
  HRAP grid
• Each computational element consists of a number
  of uniform hillslopes and conceptual channels
• Rainfall-runoff component (Sacramento model in
  the 1st version) generates fast and slow
  runoffs
• Hillslope transforms (kinematic routing) fast
  runoff into lateral channel inflow
• Channel inflow combined with slow runoff and
  upstream cell outflow is routed through a cell
  conceptual channel
• Ingests NEXRAD Stage III data
• Includes features of lumping parameters/input
  data
• Modular design to test other models

10
(No Transcript)
11
HL-RMS Structure

• Conceptualization of a grid cell

12
HL-RMS Parameterization

• A priori parameters
• Rainfall-runoff model parameter grids are
  estimated using soil/vegetation data
• Hillslope/Channel routing parameter grids, slope,
  length, area above, are calculated based on DEM
• Uniform channel shape and roughness coefficient
  is assumed at each grid cell
• Parameter adjustment
• Scaling/Replacement based on lumped or
  semi-distributed calibration of rainfall-runoff
  model
• Spatially variable channel shape and roughness
  parameters can be generated from discharge
  measurements at outlets and geomorphological
  properties at each grid cell

13
(No Transcript)
14
(No Transcript)