Flood Risk Assessment: Flood Warning and Forecasting

Question 1

Why is flood forecasting and warning important?

Answer 1

It enables timely evacuation, infrastructure protection, emergency response, and minimises damage.

Question 2

Timely warnings allow operational control like opening sluice gates or ______ floodwater to offline storage.

Answer 2

diverting

Question 3

True/False
Social media is now used to issue flood warnings in the UK.

Answer 3

True

Question 4

What defines the grade of a flood warning?

Answer 4

The predicted severity and confidence in the forecast.

Question 5

Minimum lead time required for effective warning is ___ minutes
EA targets ___ hours.

Answer 5

30 minutes
2 hours

Question 6

What three qualities must flood forecasting models have?

Answer 6

Fast, accurate, and reliable (esp. under pressure).

Question 7

Forecast models use real-time data from rainfall and ________ or level gauges.

Answer 7

streamflow

Question 8

What is the difference between flood forecasting and estimation

Answer 8

Forecasting is real-time and operational; estimation is design/planning based on return periods.

Question 9

Linear regression is most accurate when gauges are far apart.

Answer 9

False (more accurate when close, but then limited lead time)

Question 10

What is the purpose of converting regression models into matrix form?

Answer 10

To efficiently estimate parameters a and b using all time steps.

Question 11

What do you input into matrix
R in the example?

Answer 11

Columns of
uk and xk from observed data

Question 12

Parameter values (a, b) are constant during a flood event.

Answer 12

False – they may vary and need updating

Question 13

What does a Grade 1 – Flood Alert mean?

Answer 13

Flooding is possible.
Action: Be prepared.

Question 14

What does a Grade 2 – Flood Warning mean?

Answer 14

Flooding is expected.
Action: Act now.

Question 15

What does a Grade 3 – Severe Warning mean?

Answer 15

Severe flooding with danger to life.
Action: Act now, evacuate.

Question 16

What is the role of rainfall data (radar, raingauge) in flood forecasting?

Answer 16

Input data

Question 17

What is HYRAD used for in flood forecasting?

Answer 17

Rainfall visualisation

Question 18

What is the role of forecasting/modelling in flood systems?

Answer 18

Predict floods.

Question 19

What is the role of flood warning systems?

Answer 19

Dissemination and response

Question 20

IF–THEN Rules
Q: What are IF–THEN rules based on in forecasting?

Answer 20

Operator experience

Question 21

Linear Regression
What does linear regression do in forecasting?

Answer 21

Simple input–output modelling

Question 22

Transfer Function
What does a transfer function add in forecasting?

Answer 22

Storage/memory effect

Question 23

Complex Models
What do complex models represent in forecasting?

Answer 23

Full hydrologic/hydraulic systems.

Question 24

Simple regression:
x̂ₖ₊₄ = a · uₖ

Where uₖ = upstream flow at time k, and x̂ₖ₊₄ = forecast downstream flow ___ hours later.

Answer 24

4

Question 25

Formula x̂ₖ₊₁ =

Answer 25

a · uₖ₋₁ + b · xₖ This includes system memory (storage)

Question 26

A 2-hour lead time can be obtained by using:

Answer 26

x̂ₖ₊₂ = a · uₖ + b · x̂ₖ₊₁

Question 27

Formula Normal Equation θ = ___

Answer 27

(RᵀR)⁻¹ Rᵀ X̂ Where: * R: matrix of predictors * X̂: observed/target output * θ: vector of parameters

Question 28

What is the purpose of state updating?

Answer 28

To correct the state variable x̂ₖ.

Question 29

What does error updating do?

Answer 29

Adjusts for forecast error trends.

Question 30

What is parameter updating used for?

Answer 30

Re-estimating model coefficients (a, b).

Question 31

State Updating Formula x̂ₖ₊₁ = ___

Answer 31

a · uₖ₋₁ + b · xₖ

Question 32

Kalman Filter (Advanced State Updating) Kalman update equation: x̂ₖ₊₁ =____

Answer 32

a · uₖ₋₁ + b · [αₖ x̂ₖ + (1 − αₖ) xₖ]

Question 33

What does αₖ represent in the Kalman filter?

Answer 33

The weighting between predicted and observed values based on confidence.

Question 34

Error Updating Formula x̂ₖ₊₂ = ____

Answer 34

a · uₖ + b · x̂ₖ₊₁ + f(ξ) - Where ξ is the forecast error, modelled as: - ξₖ = xₖ − x̂ₖ

Question 35

Why do we use parameter updating?

Answer 35

Because the runoff process is non-linear and changing, especially during floods.

Question 36

What is the assumption of a zero order model?

Answer 36

Parameters remain unchanged

Question 37

What is the assumption of a first order model?

Answer 37

Parameters evolve with time using derivatives

Question 38

Graphical Flood Forecasting Formula x̂ₖ₊ₐ = ____

Answer 38

a · uₖ + b · xₖ₊ₐ₋₁

Question 39

What is the flood forecast line?

Answer 39

xₖ₊ₐ₋₁ ≥ (F − a · uₖ) / b

Question 40

What does statistical (radar-based) rainfall forecasting do?

Answer 40

- Extrapolates storms - Good for 1 hour

Question 41

What is the role of NWP in rainfall forecasting?

Answer 41

- Simulates physics - Good for 1–5 days

Question 42

What does hybrid rainfall forecasting do?

Answer 42

Blends both statistical and NWP using lead time weighting

Question 43

Radar Forecast Formula

Answer 43

u = a · Rᵇ Where: * u: rainfall rate * R: radar reflectivity * a, b: calibrated parameters

Question 44

What’s the difference between C-band and X-band radar?

Answer 44

- C-band: wide range, high power, immobile - X-band: lower power, mobile, better resolution

Question 45

UK NWP now uses ___ km² resolution models and ensemble forecasts (e.g. STEPS).

Answer 45

1 km²