Logo PTI Logo FedCSIS

Proceedings of the 21st Conference on Computer Science and Intelligence Systems (FedCSIS)

Annals of Computer Science and Information Systems, Volume 47

SARIMAX as a Reference Model for Emission Intensity Forecasting for HPC Systems

, , ,

DOI: http://dx.doi.org/10.15439/2026F4766

Citation: Michael Zent, , ,

Full text

Abstract. The increasing appliance of systems imitating human cognition requires reasonable, in particular transparent and resource-efficient, reference models. This work demonstrates a course of action in the field of short-term greenhouse gas emission intensity forecasting with statistical Time Series Analysis to plan HPC-system operations in advance. For that, SARIMAX modelling in combination with methods from econometrics and computer science has shown to be qualified for benchmarking by its step-by-step traceability. It also provided high-accuracy rolling day-ahead prognoses on the hourly emission intensity of the German electricity production over a span of half a year. The NMAE of 10.9\% and MASE of 73.9\% when predicting their absolute values, and NMAE of 3.4\% and MASE of 61.7\% for the corresponding hourly changes, prove the performance. Such an elaborated model, and the applied measures, can be used as a reasonable reference base to benchmark various methods in the area of time series forecasting, whether statistical ones or also those imitating human cognition.

References

  1. Y. K. Dwivedi et al., “Artificial Intelligence (AI): Multidisciplinary perspectives on emerging challenges, opportunities, and agenda for research, practice and policy,” Int. J. Inf. Manage., vol. 57, 2021. https://dx.doi.org/10.1016/j.ijinfomgt.2019.08.002
  2. L. Cozzi, “World Energy Outlook Special Report: Energy and AI,” International Energy Agency, Tech. Rep., 2025.
  3. D. Stetskov, “The Great Software Quality Collapse: How We Normalized Catastrophe,” From the Trenches, 2025. [Online]. Available: https://techtrenches.substack.com/p/the-great-software-quality-collapse
  4. I. Shumailov et al., “AI models collapse when trained on recursively generated data,” Nature, vol. 631, pp. 755–759, 2024. https://dx.doi.org/10.1038/s41586-024-07566-y
  5. P. Villalobos et al., “Will we run out of data? Limits of LLM scaling based on human-generated data,” 2024. https://dx.doi.org/10.48550/arXiv.2211.04325
  6. J. Tesch et al., “Metrology and AI: PTB’s AI Strategy,” PTB-Mitteilungen, vol. 132, no. 1, pp. 40–79, 2022. https://dx.doi.org/10.7795/310.20220199
  7. S. Makridakis et al., “Statistical and Machine Learning forecasting methods: Concerns and ways forward,” PLoS One, vol. 13, 2018. https://dx.doi.org/10.1371/journal.pone.0194889
  8. V. I. Kontopoulou et al., “A Review of ARIMA vs. Machine Learning Approaches for Time Series Forecasting in Data Driven Networks,” Future Internet, vol. 15, no. 8, 2023. https://dx.doi.org/10.3390/fi15080255
  9. N. Thompson et al., “The Computational Limits of Deep Learning,” in 9th Comput. within Limits. LIMITS, 2023. https://dx.doi.org/10.21428/bf6fb269.1f033948
  10. B. Xu and G. Yang, “Interpretability research of deep learning: A literature survey,” Inf. Fusion, vol. 115, p. 102721, 2025. https://dx.doi.org/10.1016/j.inffus.2024.102721
  11. G. E. P. Box et al., Time Series Analysis: Forecasting and Control, 5th ed. Hoboken, NJ, US: Wiley, 2015.
  12. A. Tarsitano and I. L. Amerise, “Short-term load forecasting using a two-stage SARIMAX model,” Energy, vol. 133, pp. 108–114, 2017. https://dx.doi.org/10.1016/j.energy.2018.09.157
  13. N. Elamin and M. Fukushige, “Modeling and forecasting hourly electricity demand by SARIMAX with interactions,” Energy, vol. 165, pp. 257–268, 2018. https://dx.doi.org/10.1016/j.energy.2018.09.157
  14. A. M. Karki et al., “Electric load prediction using SARIMAX,” Aust. J. Electr. Electron. Eng., vol. 22, no. 3, pp. 410–423, 2025. https://dx.doi.org/10.1080/1448837X.2024.2405438
  15. A. Kammeyer et al., “HPC operation with time-dependent cluster-wide power capping,” in Proc. 19th Conf. Comput. Sci. Intell. Syst. FedCSIS, 2024. https://dx.doi.org/10.15439/2024F1066 pp. 385–393.
  16. A. Kammeyer, “Digital Twin for Energy-Aware High-Performance Computing,” Doctoral Dissertation, Freie Universität Berlin, DE, 2025.
  17. J. Gröger et al., Umweltzeichen Blauer Engel für Rechenzentren – Hintergrundbericht zur Erarbeitung der Vergabekriterien DE-UZ 228. Dessau-Roßlau, DE: Umweltbundesamt, 2024.
  18. P. Duttilo and F. Lisi, “Short-term CO2 emissions forecasting: Insight from the Italian electricity market,” 2025. https://dx.doi.org/10.48550/arXiv.2507.12992
  19. Y. Tian et al., “Carbon dioxide emission forecast: A review of existing models and future challenges,” Sustainability, vol. 17, no. 4, 2025. https://dx.doi.org/10.3390/su17041471
  20. L. K. Gohar and K. P. Shine, “Equivalent CO2 and its use in understanding the climate effects of increased greenhouse gas concentrations,” Weather, vol. 62, no. 11, pp. 307–311, 2007. https://dx.doi.org/10.1002/wea.103
  21. B. Leiner, Spektralanalyse ökonomischer Zeitreihen, 2nd ed. Wiesbaden, DE: Gabler, 1978.
  22. R. H. Shumway and D. S. Stoffer, “Spectral Analysis and Filtering,” in Time Series Analysis and Its Applications. Cham, CH: Springer, 2025, ch. 4.
  23. W. H. Greene, “Nonstationary Processes and Unit Roots,” in Econometric Analysis. New York, NY, US: Pearson, 2018, ch. 21.2.
  24. R. J. Hyndman and G. Athanasopoulos, “Stationarity and differencing,” in Forecasting: Principles and Practice. Melbourne, AU: OTexts, 2021, ch. 9.1. [Online]. Available: https://otexts.com/fpp3/stationarity.html
  25. J. D. Hamilton, “Asymptotic Properties of a pth-Order Autoregression and the Augmented Dickey-Fuller Tests for Unit Roots,” in Time Series Analysis. NJ, US: Princeton Univ. Press, 1994, ch. 17.7.
  26. D. Kwiatkowski et al., “Testing the null hypothesis of stationarity against the alternative of a unit root: How sure are we that economic time series have a unit root?” J. Econom., vol. 54, no. 1, pp. 159–178, 1992. https://dx.doi.org/10.1016/0304-4076(92)90104-Y
  27. B. E. Hansen, “The New Econometrics of Structural Change: Dating Breaks in U.S. Labour Productivity,” J. Econ. Perspect., vol. 15, no. 4, pp. 117–128, 2001. https://dx.doi.org/10.1257/jep.15.4.117
  28. D. R. Osborn et al., “Seasonality and the Order of Integration for Consumption,” Oxford Bull. Econ. Stat., vol. 50, no. 4, pp. 361–377, 1988. https://dx.doi.org/10.1111/j.1468-0084.1988.mp50004002.x
  29. F. Canova and B. E. Hansen, “Are Seasonal Patterns Constant Over Time? A Test for Seasonal Stability,” J. Bus. Econ. Stat., vol. 13, no. 3, pp. 237–252, 1995. https://dx.doi.org/10.1080/07350015.1995.10524598
  30. R. Nau, “ARIMA models for time series forecasting,” in Statistical Forecasting: Notes on Regression and Time Series Analysis. Durham, NC, US: Fuqua School of Business, 2020, ch. 5. [Online]. Available: https://people.duke.edu/~rnau/411home.htm
  31. F. E. Harrell, “Variable Selection,” in Regression Modeling Strategies. Cham, CH: Springer, 2015, ch. 4.3.
  32. R. F. Engle et al., “Exogeneity,” Econometrica, vol. 51, no. 2, pp. 277– 304, 1983. https://dx.doi.org/10.2307/1911990
  33. K. Pearson, “Note on regression and inheritance in the case of two parents,” Proceedings of the Royal Society of London, vol. 58, no. 347352, pp. 240–242, 1895. https://dx.doi.org/10.1098/rspl.1895.0041
  34. G. J. Székely et al., “Measuring and testing dependence by correlation of distances,” Ann. Stat., vol. 35, no. 6, pp. 2769–2794, 2007. https://dx.doi.org/10.1214/009053607000000505
  35. R. J. Hyndman and G. Athanasopoulos, “Time series decomposition,” in Forecasting: Principles and Practice. Melbourne, AU: OTexts, 2021, ch. 3. [Online]. Available: https://otexts.com/fpp3/decomposition.html
  36. C. J. Willmott and K. Matsuura, “On the use of dimensioned measures of error to evaluate the performance of spatial interpolators,” Int. J. Geogr. Inf. Sci., vol. 20, no. 1, pp. 89–102, 2006. https://dx.doi.org/10.1080/13658810500286976
  37. R. J. Hyndman and A. B. Koehler, “Another look at measures of forecast accuracy,” Int. J. Forecasting, vol. 22, no. 4, pp. 679–688, 2006. https://dx.doi.org/10.1016/j.ijforecast.2006.03.001
  38. J. S. Armstrong, “Evaluating Forecasting Methods,” in Principles of Forecasting: A Handbook for Researchers and Practitioners. Dordrecht, NL: Kluwer, 2001, ch. 14.
  39. H. Akaike, “A new look at the statistical model identification,” IEEE Trans. Autom. Control, vol. 19, no. 6, pp. 716–723, 1974. https://dx.doi.org/10.1109/TAC.1974.1100705
  40. Y. Yang, “Can the strengths of AIC and BIC be shared? A conflict between model indentification and regression estimation,” Biometrika, vol. 92, no. 4, pp. 937–950, 2005. https://dx.doi.org/10.1093/biomet/92.4.937
  41. G. Schwarz, “Estimating the Dimension of a Model,” Ann. Stat., vol. 6, no. 2, pp. 461–464, 1978. https://dx.doi.org/10.1214/aos/1176344136
  42. G. M. Ljung and G. E. P. Box, “On a measure of lack of fit in time series models,” Biometrika, vol. 65, no. 2, pp. 297–303, 1978. https://dx.doi.org/10.1093/biomet/65.2.297
  43. A. C. Harvey, “Testing and model selection,” in Forecasting, structural time series models and the Kalman filter. GB: C.U.P., 1990, ch. 5.
  44. C. M. Jarque and A. K. Bera, “A Test for Normality of Observations and Regression Residuals,” Int. Stat. Rev., vol. 55, no. 2, pp. 163–172, 1987.
  45. D. Asteriou and S. G. Hall, “ARIMA Models and the Box-Jenkins Methodology,” in Applied Econometrics. London, GB: Macmillan, 2021, ch. 13.
  46. C. Chatfield, “Estimating Autocovariance and Autocorrelation Functions,” in The Analysis of Time Series: An Introduction. Boca Raton, FL, US: Chapman & Hall/CRC, 2004, ch. 4.1.
  47. T. G. Smith et al., “pmdarima API Reference: auto_arima.” [Online]. Available: https://alkaline-ml.com/pmdarima/modules/generated/ pmdarima.arima.auto_arima.html
  48. R. J. Hyndman and G. Athanasopoulos, “Transformations and adjustments,” in Forecasting: Principles and Practice. Melbourne, AU: OTexts, 2021, ch. 3.1. [Online]. Available: https://otexts.com/fpp3/ transformations.html
  49. Bundesnetzagentur (BNetzA), “SMARD-Strommarktdaten, Stromhandel und Stromerzeugung in Deutschland.” [Online]. Available: https: //www.smard.de/home/marktdaten
  50. T. G. Smith et al., “pmdarima API Reference: nsdiffs.” [Online]. Available: https://alkaline-ml.com/pmdarima/modules/generated/ pmdarima.arima.nsdiffs.html
  51. K. J. Millman et al., “Mathematical Optimization,” in Scientific Python Lectures. [Online]. Available: https://scipylectures.org/advanced/mathematical_optimization
  52. M. Stuchlik, “Cirron.” [Online]. Available: https://github.com/s7nfo/ Cirron
  53. M.-C. Düker et al., “Vector AutoRegressive Moving Average Models: A Review,” Wiley Interdiscip. Rev. Comput. Stat., vol. 17, 2025. https://dx.doi.org/10.1002/wics.70009