The article presents the results of assessing the reliability of a statistical model for developing a scenario of possible changes in the average winter air temperature over the territory of Azerbaijan. The model is based on the assumption of homogeneous soil cover in the horizontal direction, which allows the soil to be considered as a non-stationary onedimensional system with a vertical spatial coordinate and time. The statistical model used is grounded in the assumptions tha t the temporal structure of winter temperature consists of periodic and non-periodic changes over various time intervals, and the existence of a trend in the long-term dynamics of air temperature is a widely recognized and reliably established fact. Here, the random components of the time series are calculated using Schuster’s method and multiple cyclicities, and spectral analysis is conducted to identify short- and long-period fluctuations in these series. Furthermore, the expected values of air temperature residuals are computed. After this, the overall trend of the expected air temperature values was calculated as the sum of the values obtained from the linear trend and the computed residuals of air temperature. Given that the behavior of the climate system exhibits patterns of both short-period and long-period cyclicities, and their alternation within a certain time interval persists, we conducted numerical experiments based on the spectral analysis of time series of winter air temperature residuals. The calculations showed that the optimal approach is the use of a combination of cycles of 6, 9 -10, and 14-15 years. Only in the Nakhchivan Autonomous Republic was a combination of 6 and 9-10 year cycles used. To assess the quality of the proposed model, the correlation coefficient between the actual and calculated values of winter air temperature was utilized. To verify the reliability of the obtained research results using independent data (from 1998 to 2022), we compared the estimates of both current climatic changes and their expected magnitudes. The model was tested using both five -year averaged temperature changes and annual temperature changes, analyzed through graphical representations. It was found that across all physical-geographical zones and climatic periods, in 44 cases (73 %), the absolute calculation errors were ΔT ≤ 1.00C, and in 58 cases (97 %), ΔT ≤ 2.00C. These results, along with other data and the model's reliability measure, demonstrate that the proposed model adequately describes changes in winter temperature and can be used to develop a scenario for changes in average winter air temperature over the next 20-25 years.
Safarov et al. (Fri,) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: