The Challenges of Flood Monitoring across Political Boundaries: Taking Stock of Emerging Opportunities and Moving Ahead

1
The Challenges of Flood Monitoring across
Political Boundaries Taking Stock of Emerging
Opportunities and Moving Ahead

• Faisal Hossain
• Department of Civil and Environmental Engineering
• Tennessee Technological University

2
The Story of the Niger River
1. 4030 km long, 211,3200 km2
2. Flows through 5 countries
3. Drainage area comprised of 11 countries
4. Frequent river flooding induced by heavy
rainfall
Question How does one monitor early the
evolution of river flooding across political
boundaries of 5 nations, 11 administrations and a
diverse landscape?
5. Diverse climate, rainfall regime, soil
conditions, topography varying response of
landscape to rainfall
3
The Answer (Under Ordinary Circumstances)

• Acquire all necessary data to set-up a hydrologic
  model
• Static data parameters topography, soil
  classification, river network, vegetation
  (seasonal) etc.
• Dynamic data parameters rainfall,
  runoff/streamflow and soil moisture

Data Category A is relatively easy to derive from
various databases (ex Topography from Shuttle
Radar Topography Mission - SRTM Elevation data at
90m resolution).
Data Category B is the REAL CHALLENGE
space-time variability demands an accurate and
continuous REAL-TIME data streaming system
Especially Rainfall The main determinant of
the make-up of flood
4
The Problems that Magnify the Challenge of
Real-time Data Acquisition
Problem 1 11 national hydro-meteorological
agencies with no agreement for real-time data
exchange
Transboundary problems and treaties on water
resource sharing among nations is very
well-documented and researched
Problem 2 Sparse/inadequate network for data
measurement in flood-prone regions
But these treaties do not address REAL-TIME data
sharing among nations
Source Aaron Wolf, Oregon State U
5
The story continues
A 1991 United Nations survey of hydrological
monitoring networks showed "serious shortcomings"
in sub-Saharan Africa, says Rodda. "Many stations
are still there on paper," says Arthur Askew,
director of hydrology and water resources at the
World Meteorological Organization (WMO) in
Geneva, "but in reality they don't exist." Even
when they do, countries lack resources for
maintenance. Zimbabwe has two vehicles for
maintaining hydrological stations throughout the
entire country, and Zambia just has one, says
Rodda. Stokstad, E., Science, 285, 1199, 1999
Anectodal Evidence During the Mozambique floods
in 2001, there were only 4 precipitation gauges
reporting for the entire country (from Dennis
Lettenmaier, University of Washington)
6
Niger River is not alone
214 International River Basins in 1979 UN
Register 261 in 2002 (Updated)
MAURITANIA
C H I N A
SENEGAL
N E P A L
BHUTAN
MALI
CHINA
Ubiquitous in all 5 continents 145 countries are
associated in IRBs Accounts for 40 of total land
surface. gt 50 of total surface flow
I N D I A
I N D I A
MYANMAR
Source Aaron Wolf, Oregon State U
VIETNAM
LAO PDR
BANGLADESH
GUINEA
THAILAND
CAMBODIA
7
State-of-the-Art on Operational Flood Monitoring
Across Political Boundaries
Percentage within International River Basins Number of Countries
91-99 39
81-90 11
71-80 14
61-70 11
51-60 17
41-50 10
31-40 10
21-30 13
11-20 9
1-10 11
Many nations locked in IRB 30 in High Flood
Risk Zone
Basin-wide flood monitoring range depends on
knowledge of rainfall over upstream nations
Real-time rainfall data across political
boundaries not always available
15 of total death toll due to Natural Hazards is
by river flooding. 250 Billion dollar of damages
annually
Source Dr. Aaron Wolf, Oregon State Univ.
8
One solution SpaceSatellite Rainfall(?)
What we know Satellites have potential to
overcome transboundary/institutional limitations
Global Precipitation Measurement Mission in
2012
3-6 hours sampling
5X5 km - 10X10 km resolution
Coherent data calibrated to unified system
Greater accuracy expected than current products
Global coverage!
9
Other Emerging Opportunities

• Two other planned hydrologic missions can
  potentially enhance the cause of IRBs
• HyDROS (Soil Moisture) and
• WatER (Streamflow)

WaTER
HYDROS
10
Some Countries where Satellite Rainfall Could
Help (in principle)
Name of down stream country International River Basin of Total Basin Area
Cameroon Akpa/Benito/Ntem 41.8
Senegal Senegal 8.08
Ivory Coast Cavally 54.1
Benin Oueme 82.9
Botswana Okovango 50.6
Nigeria Niger 26.6
Bangladesh Ganges-Brahmaputra-Meghna 7.0
Brunei Bangau 46.0
Laos Ca/Song Koi 35.1
Cambodia Mekong 20.1
Hossain F.,and Katiyar N.(2006) EOS
Transactions 87(5)
11
However..
Recent work identified complex nature of
satellite rainfall uncertainty and non-negligible
implications at dynamic scales of surface
hydrology (TTU-UConn collaboration)1,2
How effective is the use of satellite rainfall in
large IRBs given its uncertainty?
A critical assessment of satellite rainfall to
understand the tradeoff (for current and future
scenario) needed This Can Potentially
Strengthen Communitys Argument for GPMs
importance to Society!
Foundation exists for assessment of satellite
rainfall data for a basin as a whole hydrologic
unit
BUT Transboundary limitation to modeling flow
adds a new dimension to the whole problem
1 - Hossain, F. and E.N. Anagnostou (2005). Using
a Multi-dimensional Satellite Rainfall Error
Model to Characterize Uncertainty in Soil
Moisture Fields Simulated by an Offline Land
Surface Model. Geophys. Res. Lett.vol 32. 2.
Hossain, F. and E.N. Anagnostou (2006). A Two
Dimensional Satellite Rainfall Error Model. IEEE
TGRS, vol 44(4).
Zielinski, S. (2005). Earth observation programs
may still be at risk, EOS Transactions, 86(43)
414.
12
Science Questions

• General Science Question
• How realistic is the use of satellite rainfall in
  overcoming the transboundary limitations to flood
  monitoring? Will GPM improve Flood Monitoring?
• Specific Questions
• What specific IRBs, and downstream nations would
  benefit more than others from GPM?
• What is the role played by climate and landscape?
  
• Can we develop rules of thumb for application of
  satellite rainfall data in ungauged IRBs?

13
What we need to move forward

• A global ball-park (hydrologically-relevant)
  assessment of ALL flood-prone IRBs that can serve
  as a proxy to detailed investigation on a case by
  case basis.
• Parsimonious and non-unique hydrologic modeling
  approach. (Too many IRBs So little time)
• Hydro-political component Need to
  mathematically (reasonably realistically) model
  political boundaries in IRBs. (No framework does
  that)

14
Outline of Seminar (2nd half)

• Development of a modeling approach as a way to
  initiate understanding of the impact of
  integrating satellite rainfall data for
  basin-wide flood monitoring across political
  boundaries.
• Concept Demonstration Mapping results globally.
• Broader Research and Education Plan
• A proposed long-term agenda for moving forward
  in anticipation of future hydrologic space
  missions.

15
The Hydrologic Modelling Approach
Complexity dictated by the needs of (a)
modularity (plug in and out process equations)
(b) simplicity (requiring little time to set-up
over IRBs (c) representing political boundaries
in the hydrologic framework
Geometric Configuration based on Open-Book
Watershed (first proposed by Chow and Yen at
Ilinois University (1969).
Can handle irregular DEM (if necessary)
Can handle higher order watersheds
Solves explicitly on a grid volume basis
Fully distributed (to study scales optimal data
integration)
Parsimonious in modeling geometry, topography and
surface hydrology
Katiyar, N. and Hossain, F. 2006 Env. Mod. Soft.
(In review)
16
The Grid-based Process Descriptions
Excess Rainfall Calculated using simple mass
balance (bucket) approach keeping track of soil
moisture storage
Overland Flow Excess rainfall is routed as sheet
flow from each grid along the steepest direction
Darcy-Weisbach or Mannings Equation.
River Flow Lateral inflow from overland is
routed using Mannings Equation (normal depth
equation)
Assumption for regulated rivers Dams do not act
as control structures during flooding period
(e.g. Farakka Barage in India on Ganges).
17
Excess Rainfall Calculations
s(t) soil moisture storage p(t) rainfall
qse(t) surface flow qss(t)subsurface flow
ET and GW flow can be incorporated if needed
18
Excess Rainfall Calculations
If s(t)gt Sf Sf field capacity
qss 0 if s(t) lt Sf
L grid length, ßground slope, F soil
porosity, Ks Sat. Conductivity
if s(t) gt Sb
Sb Soil storage capacityDf
D depth to bedrock (depth of soil column)
qse 0 if s(t) lt Sb
Excess rainfall Overland Routing River
Routing - Q
After Jothityankoon et al., 2002, J. Hydrol.
19
Overland Flow and River Flow
V
Darcy- Weisbach Eqn. Laminar
Mannings Eqn Turbulent
After Applied Hydrology Chow et al. 1988
20
Summary of Inputs
Static Input Geophysical Parameters
(Distributed) 1. Topography 2. Soil type (Sf,
porosity, Ks), 3. Effective Soil Depth 4. River
bed slope, 5. Channel hydraulic parameters
Minimum calibration Data available from global
databases
Dynamic Input Hydrometeorological Data
(Distributed) 1. Rainfall 2. Initial soil
moisture field (and base flow).
The Output (what you get for all
this) Streamflow at any reach of the river, flow
depth, inundation plain.
21
The Hydro-political Component
Basic Information on each riparian nation is
needed
For example Senegal River Basin Comparing the
impact of assimilating basin wide satellite
rainfall data Vs partially gauged rainfall on
streamflow modeling accuracy for Senegal How
much can Senegal benefit?

• Level of Idealization can be systematically
  reduced (for example)
• Higher ordered watersheds.
• 2. Use of in-situ DEM - political boundaries

MAURITANIA
Guinea
SENEGAL
7
Mali
35
MALI
Mauritania
Mali
50
Senegal
8
Mauritania
Senegal
GUINEA
22
Verification of the Model(The Moment of Truth
for a novice Graduate Advisor!)
Numerical Stability
Test on an Openbook of 9X6 grids at 200 m
resolution. Valley side slope 0.01 river bed
slope0.001 Mannings n 0.015
Error in Mass Balance
Discharge
?x/?t
DURATION
Physical Consistency
23
Verification (Contd.)
Time, mins
24
Concept Demonstration
Consider this scenario Govt. of a downstream
nation in an IRB wants to know (a hydrologically
relevant ball-park estimate) of improvement that
can be expected from the use of NASAs current
satellite rainfall data in improving basin-wide
flood monitoring.
NASA Satellite Rainfall Algorithm IR-3B41RT,
available real-time (hourly) on a best-effort
basis, Calibrated to TRMM Microwave data
Test Region Oklahoma Mesonet (High Quality
datasets for assessment)
Upstream
Downstream
A Hypothetical Two-Nation IRB with Open-book
configuration
25
Demonstration (Contd.)
Objective To quantify broadly the impact of
IR-3B41RT data availability over upstream
transboundary nations on downstream nations
basin-wide streamflow prediction accuracy as a
function of of ungauged transboundary area.
ASSUMPTIONS
1. Oklahoma Mesonet assumed a two-nation IRB
with a given of upstream nations area as
transboundary (i.e., ungauged).
2. Open-book configuration. An international
river assumed
3. Downstream nation has access to in-situ
rainfall data in real-time.
4. Hydrologic simulation of streamflow from
basin-wide in-situ (WSR-88D Radar) rainfall data
assumed benchmark.
26
Model Set-up

• 4 month simulation period (May Aug, 2002)
  (Typical flooding period for many nations).
• Daily time step, 0.25 degree simulation.
• 6 months spin-up performed with NOAH-LSM to
  initialize soil moisture fields.
• Geophysical parameters derived from Oklahoma
  Mesonet database, various maps, past work.
• River bed slope 0.005 Valley slopes 0.001,
  Effective soil depth 0.5 m.
• Silty loam soil, Ks0.65 cm/hr, Porosity0.5.
• Assumed river with rectangular section, 10 m
  depth, 100 m wide.

27
Results
IR-3B41RT suffers from overestimation overland
Needs Adjustment
With Bias Adjustment
28
Results (Contd.)
Improvement
Major
Minor
Negative
Impact on predicting flow with 30 intraboundary
rainfall data (70 transboundary upstream
nations)
Impact of assimilating IR-3B41RT data over
upstream nations on basin-wide streamflow
prediction uncertainty for the downstream nation
29
Speculations on IRBs
Preliminary Speculation - Setting aside ALL
assumptions
Name of down stream country International River Basin of Total Basin Area
Cameroon Akpa/Benito/Ntem 41.8
Senegal Senegal 8.08
Ivory Coast Cavally 54.1
Benin Oueme 82.9
Botswana Okovango 50.6
Nigeria Niger 26.6
Bangladesh Ganges-Brahmaputra-Meghna 7.0
Brunei Bangau 46.0
Laos Ca/Song Koi 35.1
Cambodia Mekong 20.1
Improvement
Negligible Improvement
30
More Intelligent Speculation
Based on Koppen Climate Classification
Source Encyclopedia Britannica
31
Speculation on IRBs (Contd.)
Cfa Cwa Humid Subtropical Bsh-
Semi-arid Ganges River Bangladesh (45) ? Yalu
and Tomen Rivers North Korea (20)?Limpopo
River Mozambique (35)? Senegal River
Senegal (42)?La Plata River Uruguay (45)?
32
The Next Road Ahead
Our On-going effort needs to focus on
(1) Preparing accurate input database for running
the modeling framework on major 35 IRBs in Asia,
Africa and South America.
(1) Call the Presidents/Prime Ministers of these
countries!
(2) Communicating results to Community on
Transboundary Water Resources Research to assess
hydro-political implications. (1 and 2 has
alignment with Dr. Aaron Wolf of OSU)
(3) Liasing with the greater scientific community
on WatER and GPM for synergy (alignment with Doug
Alsdorf of Ohio State U, Dennis Lettenmaier of UW
and NASA group of Bob Adler).
33
The BIG PICTURE
Broader Research Plan (on-going effort)

1. Emerging opportunities of flood monitoring
   across political boundaries using satellites.v

(2) a - Characterizing scale-dependent
complexities of satellite rainfall error
structure Space-Time Stochastic Error Model
Development SREM2D
b - Optimal integration of satellite rainfall
data in uncertain land surface models (e.g.
NOAH and CLM)
(3) Role hydrologic process complexity
identifying model complexity commensurate with
satellite rainfall data uncertainty
SREM2D
34
Towards More Insightful Monte Carlo Methods of
Uncertainty Assessment
Two different models with similar deterministic
simulation does not yield same derived
distribution
Hossain, F. 2006. ASCE J. Hydrol. Engg. (in
review)
35
BIG PICTURE (Contd.)
Broader Educational Plan (Partnership with USGS)

• The Advantages
• Modernizing Curriculum making it more exciting
  to students.
• Popularising your research at the senior and
  graduate level to identify interested candidates
  for a project.
• Bringing a student up to speed for graduate
  school (Honors program, Senior Thesis, Summer
  Internships etc.).
• Most importantly the fresh angle from the
  students thinking outside the box that I can
  not get by interacting with myself only.

• Computer-assisted instruction scheme with a
  Graphical User Interface (GUI) to improve active
  learning in classroom of
• The rainfall-runoff process (Modular Modeling
  System)
• Stochastic Theory its importance in modeling
  variability and observing implications SREM2D

Going up in Blooms Taxonomy Learning Pyramid
ASCE Body of Knowledge and ABET
36
A Proposed Research Agenda (for moving ahead
collectively)
1. Search for frameworks and metrics that
actively promote feedback between hydrologists
and satellite rainfall algorithm community.
Hydrologists need to become active users of
satellite remote sensing products.
2. Explore the role of hydrologic process
controls on flood monitoring uncertainty.
Traditional methods are less insightful. Need
wiser selection of models at ungauged basins
compatible with uncertain data.
3. Explore the relative strengths and weaknesses
of satellite-derived rainfall versus model (NWP)
rainfall Hydrologist needs to know which one to
use when, where and why.
4. Explore synergy and coordination with
communities on GPM, HyDROS and WatER and
operational flood forecasting agencies for the
ultimate Space-Borne Monitoring system someday.
Example 5 year MOU between TTU and IWM/FFWC of
Bangladesh for unhindered exchange of data and
testing schemes.
37
Acknowledgements
1. All my hard working graduate students,
especially Nitin Katiyar (and Amanda Harris and
Preethi Raj).
2. The Water Center and CEE Dept. of Tennessee
Tech for research support.
3. Aaron Wolf, collaborator and colleague, Oregon
State of University.
4. Institute of Water Modeling, and Flood
Forecasting Agency, Bangladesh.
5. Dennis Lettenmaier of UW and Doug Alsdorf of
Ohio State U. for the Big Picture Value
Discussions and Synergism.
6. George Leavesley of USGS for educational
partnership.
7. Christa Peters-Lidard and George Huffman of
NASA for their continual critique of my work.
38
Thank You!
Questions?
Map of the World if Politicians were Hydrologists