resccue Guidelines  #1


Use climate scenarios for analysing climate-related impacts in cities

What is your PROBLEM?

What climatic phenomenon affects your area?

Select the climate variables whose extreme events affect you:




Wave height

Sea level

How does it affect?





Damage to people
& infrastructure

Prediction vs Projection

Emission scenarios come from possible political decisions so, in the long term, it does not make sense to talk about climate predictions or the most likely scenario. Only in the near term, at the decadal scale or lower, the prediction term is appropriate, because from 20 years ahead all the scenarios are highly similar to each other.

Example of climate predictions (2018-2038) and climate projections (2038-2100) according to a set of model outputs and compared to observations and historical experiments (1978-2018).

What benefits would you achieve?

Highly detailed scenarios

Local/regional scenarios for the derived variables.

With high confidence level

An ensemble of “possible futures” allows to estimate the certainty level accurately.

Adapted to your situation

The simulation of the derived variables should be understandable, valuable and usable according to the users’ specific requirements.

With great quality

Simulations should also be valid and reliable, thanks to the downscaling method and the whole strategy.

Example of climate change results for a city in the period 2015 to 2100, where the center represents a zero change and the edges represent and increase of 100% with respect to the derived variables to be predicted. The line represents the middle scenario, while the shaded area represents the area of uncertainty.


This is how!

What is your problem?

Where does it affect?

  • It is a regional problem
  • It only affects a small location

How long does it last?

  • It is a very short phenomenon
  • It lasts several days

Where does it occur?

  • It is a current problem
  • It is a possible future problem

What is your time frame?

  • Days
  • Weeks
  • Months
  • Years
  • Decades

It is important to deliver high-quality information: understandable, valuable, usable, valid and reliable.

What can you use?

Where does it affect?

  • 1 day – 1 week (Weather forecast*)
  • Several weeks – 1 decade (climate anomalies PREDICTION)
  • Several decades – 1 century (climate PROJECTION)

How many models?

Minimum three, but the more the better.
The ensemble strategy allows to estimate confifidence intervals.

How many RCP scenarios?

Advisable to use at least RCP4.5 and RCP8.5**, and combine with several climate models.
Note: Emission scenarios are political wills. i.e. there IS NOT the most probable one.


Scaling spatial and temporal features using dynamical and/or statistical techniques.

What should you do?

Find historical references

Use high-quality observed data as a reference.

Obtain derived variables

se tools/models fed by climate simulations in order to obtain derived variables as flood level, water quality, etc.

Identify exceeding thresholds

Identify the hazard thresholds according to observed damages in the past and apply them to the simulated variables.

Estimate impacts

Estimate the negative effects on your sector.

Achieve confidence level

Analyse all possibilities of predictions/projections, and sort them to estimate an interval of frequency cases.

What can you use to know that?

What is your time frame?

What can you use?

What can’t you use?

One hour

Deterministic weather forecast

Deterministic climate projection (unique simulation)

One week

Probabilistic weather forecast

Deterministic climate projection (unique simulation)

A few months

Ensemble seasonal forecast

Deterministic climate projection (unique simulation)

A few years

Ensemble decadal prediction

Deterministic climate projection (unique simulation)

A few decades

Ensemble decadal prediction

Ensemble climatic projection

Deterministic climate projection (unique simulation)

Mid and late century

Ensemble climate projection

Deterministic climate projection (unique simulation)


©Resccue, 2019 · Privacy Policy