Too wet, too hot, too dry

The role of the weather persistence in Europe


Peter Hoffmann

Hydro-Climatic Risks


Content

Content

  • Topic 1: weather variability and meteorological phenomena
  • Topic 2: local weather extremes in a large-scale context
  • Topic 3: a weather-type classification for Europe and application
  • Topic 4: weather variability in a climatic context
  • Topic 5: the role of the weather persistence in Europe
  • Topic 6: re-identification of weather-types in climate scenarios
  • Topic 7: a case study
    • contextualization of extreme rainfall in Jordan




Topic 1 of 7

weather variability and meteorological phenomena

Topic 1 of 7: weather variability and meteorological phenomena

Monitoring

local temperature variability in a climatic context

Podgorica
Berlin
2022-2024

Topic 1 of 7: weather variability and meteorological phenomena

Synoptic Scale Variability

Transport of air masses across longitude and latitude
west south north
zonal meridional meridional

Topic 1 of 7: weather variability and meteorological phenomena

Zonal

high latitudes cold – low latitudes warm – moderate weather condition

Topic 1 of 7: weather variability and meteorological phenomena

Meridional – Ridge

transport across latitudes – wet, hot and dry extremes are more likely

Topic 1 of 7: weather variability and meteorological phenomena

Meridional – Trough

transport across latitudes – wet, hot and dry extremes are more likely

Topic 1 of 7: weather variability and meteorological phenomena

Omega – High over Low

atmosphere blocking – high latitude positive pressure anomalies

Topic 1 of 7: weather variability and meteorological phenomena

Blocking Index


© NOAA

🠇

Topic 1 of 7: weather variability and meteorological phenomena

Causality

linkage between circulation patterns and weather maps


training

Topic 1 of 7: weather variability and meteorological phenomena

Causality

linkage between circulation patterns and weather maps


predicting

🠇
🠇




Topic 2 of 7

local weather extremes in a large-scale context

Topic 2 of 7: local weather extremes in a large-scale context

Contextualization

meteorological phenomena and weather maps

Topic 2 of 7: local weather extremes in a large-scale context

Web Application

filtering of atmospheric fields by local timeseries

Example 1: high temperature Central Europe
http://localhost:5000/ncep?para=temp&lo=12.0&la=52.0&perc=99.00
Example 2: very high temperature Central Europe
http://localhost:5000/ncep?para=temp&lo=12.0&la=52.0&perc=99.90
Example 3: very high precipitation Central Europe
http://localhost:5000/ncep?para=temp&lo=12.0&la=52.0&perc=99.95
Example 4: very high precipitation Montenegro
http://localhost:5000/ncep?para=prate&lo=18.0&la=43 .0&perc=99.95

Topic 2 of 7: local weather extremes in a large-scale context

Heavy Rainfall Events


spatial context

Topic 2 of 7: local weather extremes in a large-scale context

Ahrtal, Germany, Flood


Geopotential Height 500hPa (Z500)

Topic 2 of 7: local weather extremes in a large-scale context

Zagora, Greece, Stormwater


Geopotential Height 500hPa (Z500)

🠇

Topic 2 of 7: local weather extremes in a large-scale context

Valencia, Spain, Stormwater


Geopotential Height 500hPa (Z500)

Topic 2 of 7: local weather extremes in a large-scale context

Weather Extremes

water vapor content – thermodynamical factor

+7%




Topic 3 of 7

a weather-type classification for Europe and application

Topic 3 of 7: a weather-type classification for Europe and application

Classification

recurring circulation patterns over Europe


https://www.dwd.de/DE/leistungen/grosswetterlage/grosswetterlage.html

30 Weather-Types
every day is asigned to one
most dominant:
WZ: West Cyclonic
TRM: Trough Central Europe
HM: High over Central Europe

Topic 3 of 7: a weather-type classification for Europe and application

Mean Local Weather Character


Potsdam

🠇
🠇

Topic 3 of 7: a weather-type classification for Europe and application

Sequences

episodes of extreme weather conditions

River Flood, Danube/Elbe, August 2002

Heatwave, Eastern Europe, September 2023





Topic 4 of 7

weather variability in a climatic context

Topic 4 of 7: weather variability in a climatic context

Decomposition

shift and spread of the distribution


Hoffmann, 2024

Topic 4 of 7: weather variability in a climatic context

Differential Warming

expected changes in the midlatitude wind systems


less zonal: strong Jet Stream

more meridional: weak Jet Stream

Topic 4 of 7: weather variability in a climatic context

long-term changes in the weather variability

zonal: west – east

  • WZ moderate weather from the North Atlantic is decreasing

meridional: north – south – north

  • increasing explained variability
  • TRM – permanent rainfall
  • SWZ – heat waves

noise:

  • low predictability

Topic 4 of 7: weather variability in a climatic context

distribution of weather persistence

🠇
🠇
🠇




Topic 5 of 7

the role of the weather persistence in Europe

Topic 5 of 7: the role of the weather persistence in Europe

Image Recognition

detection of day-to-day atmosphere similarities


Hoffmann et al., 2021

Topic 5 of 7: the role of the weather persistence in Europe

Persistence

persisting summers are hot summers in Europe


Hoffmann et al., 2021





Topic 6 of 7

re-identification of weather-types in climate scenarios

Topic 6 of 7: re-identification of weather-types in climate scenarios

Random Forest Approach

training – re-identification of weather-types in climate models
🠇
🠇

Topic 6 of 7: re-identification of weather-types in climate scenarios

Applications

in weather- and climate forecasts

  • reduction of complexity – physical fields to categories
  • detection of sequences of critical weather-types
  • early warning and future risks assessments
  • comparison of the observed and simulated weather variability
  • criteria for model ensemble evaluation and reduction
  • attribution of thermodynamic and dynamic factors in the context of climate change

Topic 6 of 7: re-identification of weather-types in climate scenarios

Challenges


Climate Explorer

🠇
🠇




Topic 7 of 7

contextualization of extreme rainfall in Jordan

Topic 7 of 7: contextualization of extreme rainfall in Jordan


Topic 7 of 7: contextualization of extreme rainfall in Jordan



UERRA

Topic 7 of 7: contextualization of extreme rainfall in Jordan


Topic 7 of 7: contextualization of extreme rainfall in Jordan


🠇
🠇
Early Warning
Rainfall Scenarios
Hazard Maps

Summary

take home messages

  • there is causal linkage between large-scale weather patterns and local meteorological phenomena and extremes
  • dynamical changes are the main source of uncertainties projecting future rainfall patterns beyond the mean temperature rise
  • there are discepancies between the real- and the model world in terms of e.g. drought conditions
  • learning from the historical data and phenomena helps to better assess regional climate scenarios
Thank you for listening!

good afternoon everybody – its a pleasure for me to give a presentation in this summer school – due to some conflicts I wasn't able to travel – doesn't matter.

my background is meteorology – my presentation is about the weather variability in the context of climate change.

in recent year Europe especially southern Europe has seen several unusual episdes of exceptional droughts, heavy rainfall and heatwaves that cannot only be explained by the temperature rise.

in the following I will present you my thoughts and relevant processes having an effect on the large-scale weather variability in Europe.

one of the main property is the weather persistence.

here is the outline of my presentation separeted in 7 subsections.

I close my presentation with some results from a case study in Jordan focusing on heavy rainfall.

the first of seven topics is about meteorological phenomena in the context of the large-scale weather variability.

the fist slide shows the weather variability in Potsdam of the past 2 years in a climatic context given as temperature anomalies with regard to 1961 to 1990.

mo matter where in Europe the chart looks similar.

the maximum range of the weather variability is in the order of 10°C.

however, the mean of the two year is +2.8°C above a mean year in the 60s to 80s.

too cold episodes are still likely.

but where do the variations come from.

local weather variability in midlatitudes is mainly determined by the transport of air masses on a large-scale across different geographical and climatic regions.

the more zonal (west to east) oriented the transport the higher the conservation of the climatic temperature contrasts between the cold pole and the warm Mediterranean – it is the most dominent circulation conditions in Europe.

meridional circulation conditions represents deviations from the zonal flow regime – the zonal transport is disturbed by a north-south component triggered by the location of high- and low pressure systems – the meridional temperature contrasts are not conserved.

lets look into detail.

the stronger the zonal (undisturbed) flow from the North-Atlantic, the faster the relocation of cyclones from west to east – rainfall patterns rain down from west to east.

that's why eastern Europe is more continental and drier than western Europe.

high latitude remain cold and low latitudes warm – the expected weather conditions are widespread unstable – the development of extremes is rather unlikely.

meridional flow conditions are part of the natural weather variability.

in case of an high pressure system over Central Europe the zonal flow is disturbed

under high pressure condition the vertical transport of air masses sink and is often associated with sunny weather condition in Central Europe.

in other parts of Europe the weather conditions are rather too wet.

a south-nord component dominates and slow down the weather variability on the local and larger scale.

Low pressure systems coming from the North Atlantic are blocked or redirected to the north or south.

existing temperature contrasts between north and south are balanced

the contrast to that describes so called trough-like patterns.

the transport of air masses is than directed from north-to-south.

cold temperature conditions in high altitude can reach far south and hit warm surface temperatures – an explosive mixture related to atmospheric stratification conditions.

surrounded by high pressure areas the relocation of the Lows and rainfall patterns is much slower than under zonal condition.

this is often associated with persisting rainfall in Central and South-East Europe.

a special feature represents omega-like circulation patterns, where high pressure prevail in high latitudes and low pressure in low latitude regions.

such a pattern tends to be very persistent and can be associated with multiple extremes at different places at the same time.

it depends on the location, who is effected.

one prominent event occured in summer 2010 over russia – simulaneously a hugh river flood ocurred in Pakistan.

there are indices to monitor the midlatidude blocking activity depends on longitude.

here are shown geopotential height anomalies given in a space-time hovemuller diagram.

march to april this year was dominated by blocking activites over the eastern part of the North-Atlantic.

the blocking high favoured the recent drought conditions in Western Europe.

there is a causal linkage between the large-scale circulation patterns (describing the transport of air masses) and the expected weather condition on the local scale.

here you can see two examples for themperature and rainfall.

both show summer conditions under high- and low-pressure conditions over Central Europe.

too hot and too sunny on the one side and too wet on the other side

now we come to topic 2

the large-scale context of local extremes

at the end, the large-scale circulation flow pattern and their recurring features determine whether it rains or not.

this is simply illustrated by this modified icon.

to better understand phenomena, we have to consider atmospheric fields at different scales.

in terms of extremes the map below represents the respective weather maps.

for demonstration I developed a web application to contextualize local extremes.

the parameter defined in URL are the location, the variable and the percentile

the raw data are daily atmopheric fields from 1981 to 2022 of the Geopotential.

the results are composite patterns of the circulation filtered by local extreme values – the curvature indicate where the air masses come from.

however, the individual context can be differ from composites.

recent years have seen several severe extreme rainfall events in parts of Europa leading to river or flash floods.

the amount of daily precipitation was partly in the order of annual values.

the locations are given here, and I hope that everybody knows it.

firstly the event in Germany in July 2021.

second, the severe rainfall in Greece in early September 2023 – within this event the amount of daily rainfall was in the order of annual totals – later a medicane has formed in the eastern Mediterranean and destroyed parts of the coast of Libya.

finally Valencia, where very cold air masses were laying over warm near surface – and this for days.

clear is, critical weather patterns can trigger local extremes.

most of these are meridional weather-types.

the direkt link to climate change is the temperature rise and the water vapor content.

the recent EU copernicus program reported an increase by about 7%

a warmer atmosphere can storage mor watervapor and potential rainwater like a sponge

this makes extreme droughts and extreme rainfall more likely.

local thunderstorm conditions have a higher potentials for stormwater.

it depends on the topography, whether the rainwater can accumulate dangerously.

in the followin, we can bring order in the complex features using classified weather-types.

here you can see a diversity of recurring circulation patterns over Europe – it based on an synoptical classification of weather-types by experts.

shown are composite patterns of the respective weather-types.

every day is assigned to one of these 30 acronyms

the upper row represents the most dominant ones explaining about 50% of the total weather variability.

WZ is characterized by a strong gradient between North and South nea– westerly winds from the North-Atlantic dominates.

there are other prominent features like TRM, TM, HM, SWZ, TB, HNZ.

on the local scale each weather-types is seasonally associated by mean weather characteristics

here is given a fully matrix of temperature and precipitation for Potsdam.

you can identify which is weather-type is retrospectively associated with warm/cold wet/dry conditions.

every location in Europe has another assignments – dry in Central Europe can be wet in Eastern Mediterranean.

the development of extreme weather espisodes is also visible in the sequence of critical weather-types.

here are two examples for a flood events and warm period.

you can see a persistence of one or similar weather-types for one or two weeks

weather and climate belongs together – climate is the distribution of weather but how it can be effected.

this will try to explain by this distribution function for weather and climate, separetely.

usually, both are merged and not separated.

one distribution is shifted to higher values and the other distribution is compressed leading to a larger spread on both ends of the distribution.

it impressively show the effect of climate change on the weather variability and extremes.

such an attribution is enabled by the classification of weather-types.

changes in the shape of the distribution is an indicator for dynamical changes – the temperature rise only effects the shift

what are the mechanisms or possible processes behind.

the main processes are the differential warming and arctic amplification – land/ocean and high/low latitudes.

the stronger the amplified warming of the arctic and the landmasses the weaker the midlatitude Jet Stream.

a weak Jet Stream tend to meander more often in north-south direction.

this phenomena can amplify the probability of persistent weather and extreme conditions prefered in boreal summer.

long-term trend analyses of European weather-types show decline of the moderate west-wind weather-type from the past to the present period.

the missing is compensated by new dominant and meridional weather-types TRM, SWZ. – both favour extreme weather conditions

to summarize the weather variability is nowadays explained by new dominate weather-types

now, we come to the role of the weather persistence in Europe.

how did we do that?

we applied an estabished image recognition approach and compared the similarity of consequtive days.

the similarity is high, if the iso-contours of the atmospheric fields run in similar paths over days.

the more chaotic the lower the similarity and the persistence.

the application to climate models was sobering, because we could not find a feature like this in the historical period.

in 2021 we published a study about atmosphere simularities in the day-to-day weather variability.

for summer condition an increase in the similarity and liknk this with an increase of the weather persistence.

the effected regions are the North-Atlantic, Europe and Siberia.

additionally we correlated this with mean temperature and precipitation and found over Europe positive values for temperature and mostly negative values for precipitation.

persistent summers in Europe are mostly hot and dry summers.

in topic 6 I will show how to use machine learning algorithms for a re-identification of existing weather-types in weather and climate forecast.

weather-types are implifications of recurring atmospheric fields that can be support experts for interpretion.

used is a random forest approach to train the relation between daily atmopheric fields over Europe and daily weather-types of an existing classification for the historical period

the method can be evaluated by slitting the historical period into a training and a predicting period.

finally we apply the derived rules to predict and convert the unclassified atmospheric field from model to a sequence of daily weather-types.

for proving, we calculate composite maps for each weather-type and compare these with the observations.

based on this sequence we can easily compare long-term trends or persistence.

there are further applications, such as ...

one challenge is shown here.

one average historical simulations regional climate model show an increase in annual precipitation in most part of Europe.

however, the observed pattern show more complex features.

these are likely liked to change in the weather-types.

in the last part, I would like to show some results from another finished project that dealt with heavy rainfall in the eastern Mediterranean.

the focus region was Jordanien, a country in the Near-East Region.

Jordan is a country located in a aride climate zone. The potential evaporation is higher than the amount of rainfall.

our focus regions were the capital Amman and famous rock city Petra.

in total I visited Jordan 4 times.

<span style="color:magenta;float:right;font-size:14pt">GSMap</span>

in summer there no rain and in winter heavy rainfall events can trigger severe flash floods.

one of the most severe in Amman ocurred in February, 2019. The downtown of Amman was completely flooded.

the total amount of rainfall was in ther order of 100 mm within several hours.

one part was a retrospective analysis of historical heavy rainfall events.

the main driver are critical weather maps showing a trough-like pattern over the eastern Mediterranean.

low pressure systems can trigger intense rainfall in combination with too warm sea surface temperature.

Finally, I would like to thank you for your attention and hope that I have been able to provide you with some useful information.

f you have any questions, please feel free to contact me afterwards.

## Weather Variability <hr> ##### **Too wet, too hot, too dry – The role of the weather persistence** ![w:150 h:150](./svg/WIND.svg) ![w:150 h:150](./svg/UMBRELLA.svg) ![w:150 h:150](./svg/TSTORM.svg) ![w:150 h:150](./svg/TORNADO.svg) ![w:150 h:150](./svg/THERMOMETER.svg) ![w:150 h:150](./svg/SUN.svg) ![w:150 h:150](./svg/SNOW263.svg) ![w:150 h:150](./svg/smoke.svg) ![w:150 h:150](./svg/HURRICANE.svg) ![w:150 h:150](./svg/DRIZZLE.svg) ![w:150 h:150](./svg/COMPASS.svg)