Modeliranje in napovedovanje emisij CO2 za Slovenijo in Evropo z metodami analize časovnih vrst

Vidovič, Blaž

| | SLO | ENG | Cookies and privacy

Bigger font | Smaller font

First page > Show document

Show document

Title:	Modeliranje in napovedovanje emisij CO2 za Slovenijo in Evropo z metodami analize časovnih vrst
Authors:	ID Vidovič, Blaž (Author) ID Bogataj, Miloš (Mentor) More about this mentor... ID Potrč, Sanja (Comentor)
Files:	MAG_Vidovic_Blaz_2026.pdf (4,89 MB) MD5: A9D8124CC0E64330398CCCC341A81323
Language:	Slovenian
Work type:	Master's thesis/paper
Typology:	2.09 - Master's Thesis
Organization:	FKKT - Faculty of Chemistry and Chemical Engineering
Abstract:	V magistrskem delu obravnavamo napovedovanje emisij fosilnega CO2 za Slovenijo in izbrane evropske države do leta 2050 z modeli časovnih vrst. Namen dela je razviti in validirati napovedne modele ter oceniti vrzel med napovedanimi trendi in podnebnimi cilji. Na podlagi podatkov zbirke Global Carbon Budget za obdobje 1990–2024 smo v okolju MATLAB razvili univariatne modele ARIMA, modele ARIMAX z eksogenim prelomljenim trendom in nadomestno spremenljivko za pandemijo COVID-19 ter večrazsežni model VAR. Vsi modeli sledijo enotnemu postopku, ki vključuje preverjanje stacionarnosti, izbiro po informacijskih kriterijih, diagnostiko ostankov, izvenvzorčno validacijo in napoved z intervali zaupanja; za Slovenijo je dodatno obravnavan tudi scenarij zaprtja Termoelektrarne Šoštanj. Rezultati kažejo, da modeli omogočajo izvedljive napovedi, a z velikimi intervali zaupanja zaradi kratkega podatkovnega obdobja in dolgega napovednega horizonta. Ugotovili smo, da vključitev dodatne modelne strukture, predvsem determinističnega prelomnega trenda, izboljša prileganje. Za Slovenijo je model ARIMAX s prelomljenim trendom dosegel najnižjo napovedno napako, model VAR pa razkrije, da večino napovedne variance emisij pojasnjujejo njihovi lastni šoki ter da emisije in gospodarska rast dolgoročno kažeta različni smeri gibanja. Pri evropskih državah z izrazito padajočimi emisijskimi trendi lahko neomejena ekstrapolacija povzroči fizikalno nesmiselne, tudi negativne vrednosti emisij. Primerjava napovedanih emisijskih trajektorij s podnebnimi cilji kaže, da nadaljevanje zgodovinskih trendov zmanjševanja emisij pri več obravnavanih primerih ne bi zadostovalo za dosego zastavljenih ciljev. Rezultati zato poudarjajo potrebo po hitrejšem zmanjševanju emisij in, kjer je to predvideno v podnebnih strategijah, tudi po dopolnilnih ukrepih za povečevanje ponorov oziroma odstranjevanje CO2.
Keywords:	napovedovanje emisij CO2, časovne vrste, ARIMA, ARIMAX, vektorska avtoregresija
Place of publishing:	Maribor
Year of publishing:	2026
PID:	20.500.12556/DKUM-99050
Publication date in DKUM:	09.09.2026
Views:	203
Downloads:	10
Metadata:
Categories:	KTFMB - FKKT
:	Copy citation

Average score:	(0 votes)
Your score:	Voting is allowed only for logged in users.
Share:

Hover the mouse pointer over a document title to show the abstract or click on the title to get all document metadata.

Licences

License:	CC BY-NC-ND 4.0, Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International

Link:	http://creativecommons.org/licenses/by-nc-nd/4.0/
Description:	The most restrictive Creative Commons license. This only allows people to download and share the work for no commercial gain and for no other purposes.
Licensing start date:	28.07.2026

Secondary language

Language:	English
Title:	Modelling and Forecasting CO2 Emissions for Slovenia and Europe Using Time Series Analysis Methods
Abstract:	In this master's thesis, we address the forecasting of fossil CO2 emissions for Slovenia and selected European countries up to 2050 using time series models. The aim of the work is to develop and validate forecasting models and to assess the gap between the projected trends and the climate targets. Based on Global Carbon Budget data for the period 1990–2024, we developed univariate ARIMA models, ARIMAX models with an exogenous broken trend and a dummy variable for the COVID-19 pandemic, and a multivariate VAR model in MATLAB. All models follow a common procedure that includes stationarity testing, model selection by information criteria, residual diagnostics, out-of-sample validation, and forecasting with confidence intervals; for Slovenia, a scenario of the closure of the Šoštanj Thermal Power Plant is additionally considered. The results show that the models produce feasible forecasts, but with wide confidence intervals due to the short data period and the long forecast horizon. We found that including additional model structure, particularly a deterministic broken trend, improves the fit. For Slovenia, the ARIMAX model with a broken trend achieved the lowest forecast error, while the VAR model indicates that most of the forecast variance of emissions is explained by their own shocks and that emissions and economic growth move in different directions in the long run. For European countries with strongly declining emission trends, unconstrained extrapolation can produce physically meaningless, even negative, emission values. A comparison of the projected emission trajectories with climate targets shows that a continuation of historical emission-reduction trends would, in several of the cases examined, be insufficient to meet the stated targets. The results therefore underscore the need for more rapid emission reductions and, where envisaged in climate strategies, for complementary measures to enhance sinks or remove CO2.
Keywords:	CO2 emissions forecasting, time series, ARIMA, ARIMAX, vector autoregression

Comments

Leave comment

You must log in to leave a comment.

Comments (0)

0 - 0 / 0

There are no comments!

Back