Skip to main content

LUP Student Papers

LUND UNIVERSITY LIBRARIES

A Sampling-Based Approach to Estimate Tree-Ring Width Chronologies

Carlsen, Mads LU (2025) In Master’s Theses in Mathematical Sciences BERM08 20251
Mathematics (Faculty of Sciences)
Abstract
In the field of dendrochronology, scientists work with the process of ordering different trees
in time by using their annual tree ring width (TRW). The TRW data are logged in time
series and cross-dated with each other to find out in which period of time they both were
alive. The two TRW series are then conjoined by taking the average of their ring widths
in each overlapping year, and this can be repeated adding more and more trees to create a
chronology. This is done adding one tree after the other, analysing where each individual
tree ’fits in’, adding the ’best fitting’ series, then next best, third best and so forth. This
is a time-consuming process, matching TRW series and determining which series fit in and
which do not. In... (More)
In the field of dendrochronology, scientists work with the process of ordering different trees
in time by using their annual tree ring width (TRW). The TRW data are logged in time
series and cross-dated with each other to find out in which period of time they both were
alive. The two TRW series are then conjoined by taking the average of their ring widths
in each overlapping year, and this can be repeated adding more and more trees to create a
chronology. This is done adding one tree after the other, analysing where each individual
tree ’fits in’, adding the ’best fitting’ series, then next best, third best and so forth. This
is a time-consuming process, matching TRW series and determining which series fit in and
which do not. In this project a workflow was developed with the aim of automatically cross-
dating a given number of TRW series in a probabilistic manner, utilising principles from
different statistical methods including Markov Chain Monte Carlo (MCMC) and Artificial
Bee Colony (ABC). The workflow was developed to sample from the multivariate distribu-
tion of the TRW series’ datings, and gradually converge towards the distribution showing
the TRW series’ most probable datings. The developed method proved to be somewhat
successful in correctly cross-dating both artificially generated and real tree ring data. It
furthermore managed to provide insight into the uncertainty of individual trees’ ring width
data. Although this workflow is an early-stage approach and possible improvements were
identified, it could potentially help scientists develop chronologies faster and provide them
with uncertainty information, whilst merely acting as a tool and not a replacement of the
scientist. (Less)
Popular Abstract (Swedish)
När de flesta ser eller hör talas om träds årsringar, är deras första tanke troligen att man
uppskatta ett träds ålder genom att räknar ringarna. Om man skulle analysera ringarna
mer noggrant och mäter deras tjocklek, kan de bidra till att reda ut mysterier kring tidigare
klimat och växtförhållande. Det intressanta med trädringar är, att de växer tjockare under
varma och gynnsamma förhållanden än under kalla. Det innebär att om man hade ett träd så
tjockt och gammalt att det innehöll hundra- eller tusentals årsringar, skulle man kunna säga
något om hur mycket värmare eller kallare klimat var tidigare jämfört med idag. Trotts de
allra flesta träd inte lever så länge än, är det dock möjligt att imitera detta fenomen genom
att använda... (More)
När de flesta ser eller hör talas om träds årsringar, är deras första tanke troligen att man
uppskatta ett träds ålder genom att räknar ringarna. Om man skulle analysera ringarna
mer noggrant och mäter deras tjocklek, kan de bidra till att reda ut mysterier kring tidigare
klimat och växtförhållande. Det intressanta med trädringar är, att de växer tjockare under
varma och gynnsamma förhållanden än under kalla. Det innebär att om man hade ett träd så
tjockt och gammalt att det innehöll hundra- eller tusentals årsringar, skulle man kunna säga
något om hur mycket värmare eller kallare klimat var tidigare jämfört med idag. Trotts de
allra flesta träd inte lever så länge än, är det dock möjligt att imitera detta fenomen genom
att använda flera träd från olika tidsperioder. I praktiken görs detta genom att noggrant
mäta tjockleken av varje årsring från flera träd (årsringstjocklek förkortas ofta till TRW efter
det engelska uttrycket ’Tree Ring Width’), varefter varje träd får sin TRW-data loggad som
en tidsserie. Dessa tidsserier jämförs därefter med parvis för att ta reda på om och hur de
’passar ihop’, det vill säga där mönster av toppar och dalar överensstämmer. Denna process
kallas för korsdatering, eftersom man bestämmer trädens relativa ’dateringar’, det vill säga
vilket år en träds TRW-tidsserie börjar och slutar i förhållande till en annans. Fler och fler
träd kan sedan kopplas samman i denna kronologi, vilket resulterar i en lång tidsserie, som
om man hade haft ett mycket gammalt träd.
Denna korsdateringsprocess är dock tidskrävande, även för experter, och osäkerhets-
bedömning mellan individuella årsringsserier är också del av processen. Detta ger upphov
till utvecklingen av ett verktyg som kan hjälpa forskare genom att föreslå sannolika da-
teringar av varje träds TRW-serie. Dessutom skulle verktyget samtidigt identifiera vilka
träd, om några, som inte ’passar in’ med de andra. Detta är precis vad det här projektet
försöker att lösa. Lösningen kommer att använda statistiska metod och principer från Arti-
ficial Bee Colony (artificiell bikoloni) och Markov Chain Monte Carlo-sampling. Produkten
resulterande från detta verktyg kommer att vara av olika typer, beroende på användarens
preferens. Möjligheterna inkluderar visualiseringar av träd-dateringarnas marginalfördel-
ningar samt grafer som visar TRW-tidsserierna utifrån deras respektive dateringar. Detta
skulle potentiellt kunna hjälpa forskare genom att automatisk utvärdera träds relative da-
teringar, och ge en jämförelsegrund för deras egna resultat. Detta verktyg är därmed tänkt
som et hjälpmedel för forskare och kommer inte på något sätt att ersätta dem. (Less)
Popular Abstract (Danish)
Når de fleste enten ser eller hører om træers årringe, er deres første tanke sandsynligvis, at
man kan vurdere et træs alder ved at tælle dets ringe. Analyserer man ringene mere nøje
og måler deres tykkelse, kan de være med til at opklare mysterier om fortidens klima og
vækstforhold. Det interessante ved træers ringe er, at de i varmere og mere gunstigt vejr
vokser sig tykkere end de gør i koldere vejr. Det vil sige, at hvis man havde et træ så tykt
og gammelt at det havde hundrede- eller tusindvis af årringe, ville man kunne sige noget om
hvor meget koldere eller varmere fortidens klima var sammenlignet med i dag. Selvom langt
de fleste træer ikke lever nær så lang tid, er det stadig muligt at efterligne dette fænomen
ved at bruge... (More)
Når de fleste enten ser eller hører om træers årringe, er deres første tanke sandsynligvis, at
man kan vurdere et træs alder ved at tælle dets ringe. Analyserer man ringene mere nøje
og måler deres tykkelse, kan de være med til at opklare mysterier om fortidens klima og
vækstforhold. Det interessante ved træers ringe er, at de i varmere og mere gunstigt vejr
vokser sig tykkere end de gør i koldere vejr. Det vil sige, at hvis man havde et træ så tykt
og gammelt at det havde hundrede- eller tusindvis af årringe, ville man kunne sige noget om
hvor meget koldere eller varmere fortidens klima var sammenlignet med i dag. Selvom langt
de fleste træer ikke lever nær så lang tid, er det stadig muligt at efterligne dette fænomen
ved at bruge flere træer fra forskellige tidsaldre. I praksis gøres dette ved at nøje udmåle
tykkelsen af hver årring fra adskillige træer (årringetykkelse forkortes ofte til TRW efter det
engelske udtryk ’Tree Ring Width’), hvorefter hvert træs TRW-data logges i en tidsserie.
Disse tidsserier sammenlignes derefter parvist for at undersøge om og hvordan de ’passer
sammen’, det vil sige hvor mønstre som toppe og dale stemmer overens. Denne proces
kaldes for krydsdatering, eftersom man fastslår træernes relative ’dateringer’, altså hvilket
år et træs TRW-tidsserie starter og slutter i forhold til et andet træs. Flere og flere træer
kan derved kobles sammen i denne såkaldte kronologi, hvilket resulterer i én lang tidsserie,
som havde man med et gammelt træ at gøre.
Denne krydsdateringsproces er dog tidskrævende, selv for eksperter, og vurdering af
usikkerheder mellem individuelle årringeserier er også en del af denne tidskrævende proces.
Dette giver anledning til udviklingen af et værktøj, der kan hjælpe forskere ved at foreslå
sandsynlige dateringer af hvert træs TRW-serie. Derudover ville værktøjet samtidig udpege
hvilke, om nogle, træer, der ikke ’passer ind’ med de andre. Denne opgave er netop, hvad
dette projekt forsøger at løse. Løsningen vil gøre brug af statistiske metoder og principper
fra Artificial Bee Colony (kunstig bikoloni) og Markov Chain Monte Carlo-prøvetagning.
Produktet resulterende fra dette værktøj vil være af forskellige typer, afhængigt af brugerens
præferencer. Mulighederne vil omfatte visualisering af træ-datiringenes marginalfordelinger
og grafer, som viser TRW-tidsserierne givet deres respektive dateringer. Dette ville potentielt
kunne hjælpe forskere, ved automatisk at kunne vurdere træers relative dateringer, og give
et sammenligningsgrundlag til deres egne resultater. Dette værktøj er dermed ment som en
hjælp til forskere, og vil altså langt fra kunne erstatte dem. (Less)
Popular Abstract
When most people see or hear about tree rings, they likely think of the fact that a tree’s
age can be estimated by counting the number of rings visible on its trunk when cut down.
However, when one analyses these rings a bit more carefully and measures their thickness,
they can help solve mysteries about past climate and growth conditions. What’s particularly
interesting about tree rings is that they grow thicker during warm years with favourable
conditions, and remain relatively thin during colder and less favourable years. This means
that if one had a tree so thick that it would have hundreds or even thousands of rings, one
could make estimates of how much colder or warmer the past was compared to today. Even
though most trees do... (More)
When most people see or hear about tree rings, they likely think of the fact that a tree’s
age can be estimated by counting the number of rings visible on its trunk when cut down.
However, when one analyses these rings a bit more carefully and measures their thickness,
they can help solve mysteries about past climate and growth conditions. What’s particularly
interesting about tree rings is that they grow thicker during warm years with favourable
conditions, and remain relatively thin during colder and less favourable years. This means
that if one had a tree so thick that it would have hundreds or even thousands of rings, one
could make estimates of how much colder or warmer the past was compared to today. Even
though most trees do not live nearly that long, it is still possible to mimic this phenomenon by
using multiple trees that lived during different time periods. In practice, scientists carefully
measure the width of each ring in multiple tree samples and record these as time series.
They then compare these time series in pairs to determine whether and where they ‘fit’
together, that is, where peaks and valleys align. This process is called cross-dating, as it
involves determining the trees’relative ‘dates’, i.e., the years in which each tree’s TRW
time series begins and ends in relation to others. One can then gradually add more and
more trees to this time series, building it further and further back in time, resulting in a
time series (known as a chronology) equivalent to having a sample from one very old tree.
However, the process of cross-dating many trees’ TRW series is time-consuming, even for
experienced scientists, and addressing the uncertainty of individual TRW series is also part
of this time-consuming process. This presents an opportunity to develop a tool that can
assist scientists by suggesting possible datings of each TRW series. It would further assist
by pointing out which TRW series appear to fit together, and which, if any, do not. This is
precisely the challenge that this project is based on and seeks to address. Principles will be
used from different statistical approaches like Artificial Bee Colony and Markov Chain Monte
Carlo sampling. The resulting output of such a tool will be in different forms, depending on
the user’s preferences. These include visualisations of the marginal distributions of individual
trees’estimated dates and graphical plots of their TRW time series at those dates. This
may aid scientists by providing estimated relative dates between trees, along with an output
to compare their own results against. This tool is thereby meant to assist the scientists and
not replace them. (Less)
Please use this url to cite or link to this publication:
author
Carlsen, Mads LU
supervisor
organization
alternative title
En Samplings-Baserad Metod för att Skatta Kronologier för Årsringsbredd
course
BERM08 20251
year
type
H2 - Master's Degree (Two Years)
subject
keywords
Chronology, cross-dating, dendrochronology, sampling, time series, tree rings, TRW series, uncertainty quantification
publication/series
Master’s Theses in Mathematical Sciences
report number
LUNFTB-3003-2025
ISSN
1404-6342
other publication id
2025:E59
language
English
id
9208797
date added to LUP
2026-06-09 13:51:29
date last changed
2026-06-10 15:15:51
@misc{9208797,
  abstract     = {{In the field of dendrochronology, scientists work with the process of ordering different trees
in time by using their annual tree ring width (TRW). The TRW data are logged in time
series and cross-dated with each other to find out in which period of time they both were
alive. The two TRW series are then conjoined by taking the average of their ring widths
in each overlapping year, and this can be repeated adding more and more trees to create a
chronology. This is done adding one tree after the other, analysing where each individual
tree ’fits in’, adding the ’best fitting’ series, then next best, third best and so forth. This
is a time-consuming process, matching TRW series and determining which series fit in and
which do not. In this project a workflow was developed with the aim of automatically cross-
dating a given number of TRW series in a probabilistic manner, utilising principles from
different statistical methods including Markov Chain Monte Carlo (MCMC) and Artificial
Bee Colony (ABC). The workflow was developed to sample from the multivariate distribu-
tion of the TRW series’ datings, and gradually converge towards the distribution showing
the TRW series’ most probable datings. The developed method proved to be somewhat
successful in correctly cross-dating both artificially generated and real tree ring data. It
furthermore managed to provide insight into the uncertainty of individual trees’ ring width
data. Although this workflow is an early-stage approach and possible improvements were
identified, it could potentially help scientists develop chronologies faster and provide them
with uncertainty information, whilst merely acting as a tool and not a replacement of the
scientist.}},
  author       = {{Carlsen, Mads}},
  issn         = {{1404-6342}},
  language     = {{eng}},
  note         = {{Student Paper}},
  series       = {{Master’s Theses in Mathematical Sciences}},
  title        = {{A Sampling-Based Approach to Estimate Tree-Ring Width Chronologies}},
  year         = {{2025}},
}