Impact of the Sun on the Maximum Sustained Wind Speed at the Rancho Boyeros Meteorological Station, Cuba, Using ROR Modeling

The ability to predict climate variables in advance offers the possibility of being able to act in time and reduce adverse impacts, that is, adapt to the effects of climate change and variability. Increased preparedness for extreme climate events contributes significantly to reducing vulnerability (IPCC, 2007b) [1]. Climate prediction is one that predicts the average climate conditions for periods of duration from one month to one or two years. In practice, two large groups are distinguished: those who make forecasts of the value of the element in question, deterministic forecasts, and those who forecast the probability of occurrence of a certain value of the element, probabilistic forecasts. This work uses deterministic statistical models using the regressive objective regression.

In the most recent Scientific Assessment report of the Intergovernmental Panel on Climate Change (IPCC, 2007 a) [2] it is concluded that warming is unequivocal, this will bring disruptions in other climatic variables such as precipitation or rainfall and also in the behavior of winds and hurricanes. Weather and climate forecasts are an important element in the life of modern society. Having a forecasting system on several scales (monthly, daily and annual) allows us to have a powerful tool in planning activities of a social economic nature.

In Cuba, important work has been carried out to determine among groups of primary and calculated predictors of dynamic type and of the Temperature-Humidity complex the potential future predictors that intervene in the selection of real predictors for the rain forecast in Cuba [3], regarding hurricanes, important models have been obtained using regression [4], in the area Gray et al [5] have obtained good results regarding error levels. In the present work, a pure statistical forecast is used, searching in previous steps (Lags) for the informativeness of the process to be modeled, in our case the tri-hourly atmospheric pressure for CUBA and the possible impact on the weather of the hurricanes that will appear in the area. from the Caribbean. The objective of this work is to model and forecast the speed of the sustained wind by measuring the impact of the Sun’s magnetic activity index on it, establish if there is any trend in this variable and see which are the main statisticians of the models, with the help of the Objective Regressive modeling ROR [6-11].

To carry out this work, we had a climatic database from the Rancho Boyeros station, Cuba corresponding to the period from January 1973 to December 2010 of monthly data, the ROR methodology was used for the modeling and forecasting of the variable sustained wind (VM), in addition to the data of three variables that represent the impact of the Sun, VS1, VS2 and VS3, which correspond to sunspots, the aa index of magnetic activity of the Sun and the neutron flux called as KN Kiel neutron respectively.

In Table 1, you can see the most significant statistics of the variables studied, particularly VM is highlighted in red with an average value of 25.96 km/h (Figure 1) with a standard deviation of 4.83, the maximum value reached corresponds to 51.2. Km/h and the minimum of 15.2 km/h.

Table 1:Descriptive statistics.

First, the VM was modeled using the ROR methodology, this short-term model explains 98.7 of the variance with an error of 4.26 km/h, the Durbin Watson statistic is small so the model is open to the inclusion of more variables, however, with what is obtained, short-term behavior can be predicted (Table 2).

Table 2:Summary of the model^c,d.

^a. Predictors: Step118, Step119, NoC, DS, DI, Lag4VS2
^b. For regression through the origin (the model without intercept), R square measures the proportion of the variability in the dependent variable about the origin explained by the regression. This CANNOT be compared to the R squared for models that include intercept.
^c. Dependent variable: MAXIMUM SUSTAINED SPEED
^d. Linear regression through the origin
Fisher’s F is 2695.327, significant at 100%.

The model in question can be seen in Table 3. It depends on DS and DI, which are variables of the ROR methodology representing Sawtooth and inverted Sawtooth, NoC, is the trend of the series which is negative, which is which indicates a decrease over time of VM, Lag4VS2, represents the impact of the Sun 4 months ago, as VS2 increases, VM decreases, the Step variables are the impact of case 119 and 118 in the series, both statistically significant. The impact 4 months ago coincides with other works where as the minimum temperature increases, four months ago the number of admissions for cerebrovascular diseases increases [11], this indicates the Sun as also a regulator of temperature and cerebrovascular diseases. Everything points to the existence of a small 4-month cycle caused by THE Sun and that impacts climate phenomena, health and probably the economy and society, coinciding with the 4 months of change of the seasons of the year.

In Figure 2 you can see the value of VM and its predicted value.
We wanted to improve the model so we added other variables obtaining a model with a lower error of 3.04 and a higher explained variance of 99.3%. This time we use lag 88 from VS2 and lag 1 month ago from VM, the Durbin Watson statistic this time is 2.08 which tells us that it is not necessary to include more variables (Table 4).

Table 3:Coeficients^a,b.

^a. Dependent variable: MAXIMUM SUSTAINED SPEED
^b. Linear regression through the origin

Table 4:Coeficients^a,b.

^a. Dependent variable: MAXIMUM SUSTAINED SPEED
^b. Linear regression through the origin

This time the predicted behavior is better (Figure 3).

Finally, we carried out a long-term model using the 22-year cycle of the Sun, the results can be seen in Table 5. The estimation error decreased even though the Durbin Watson also decreased.

Table 5:Summary of the model^c,d.

a. Predictors: Lag220VS2, DS, DI, Lag264VM, NoC
b. For regression through the origin (the model without intercept), R square measures the proportion of the variability in the dependent variable about the origin explained by the regression. This CANNOT be compared to the R squared for models that include intercept.
c. Dependent variable: MAXIMUM SUSTAINED SPEED
d. Linear regression through the origin

The forward forecast is shown in Figure 4, an increase in VM is seen.

Table 6:Coeficients^a,b.

^a. Dependent variable: MAXIMUM SUSTAINED SPEED(VM)
^b. Linear regression through the origin

The short and long-term models of the Maximum Sustained Speed were obtained, determining the modulating impact of the Sun through the index of magnetic activity of the Sun, which is a geomagnetic index, the short-term trend is to decrease VM, while at in the long term the trend is increasing, the main statisticians of the models were shown. Everything points to the existence of a small 4-month cycle caused by THE Sun and that impacts climate phenomena, health and probably the economy and society.

We would like to thank Doctor of Science Uvedel Bernabe del Pino for kindly providing us with the data without which this work would not have been possible, as well as his teachings throughout our training as a researcher and Dr. Rigoberto Fimia Duarte for his unconditional help in our investigations.

• IPCC (2007b) Summary for Policymakers (Resumen para decisores). En: Climate Change 2007: Impacts, Adaptation and Vulnerability. Contribution of Working Group II to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change. (ML Parry, OF Canziani, JP Palutikof, PJ van der Linden and CE Hanson, eds.), Cambridge University Press, Cambridge, UK, pp. 7-22.
• IPCC (2007 a) Summary for Policymakers. Climate Change 2007: Mitigation. Contibution of Working Group III to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change (B Metz, OR Davidson, PR Boshc, R Dave, LA Meyer, eds.,), Cambridge University Press, Cambridge, United Kingdom and New York, NY, USA, pp. 24.
• Llanes MT (2010) Informatividad y campos medios de los predictores para el pronóstico de la lluvia en tres estaciones meteorológicas de Cuba. Revista Cubana de Meteorología 16(1): 2010.
• Ballester M, Cecilia González Pedroso, Ramón Pérez Suárez (2004) Modelo estadístico para el pronóstico de la actividad ciclónica en el océano atlántico, el golfo de México y el mar Caribe. Revista Cubana de Meteorología 11(1).
• Gray WM, CW Landsea and P Klotzbach (2003) Seasonal Hurricane Forecasting, April 2003.
• Oses R, Grau R (2011) Modelación regresiva (ROR), versus modelación ARIMA, usando variables dicotómicas en mutaciones del VIH. Universidad Central Marta Abreu de las Villas, 25 de Febrero. Editorial Feijó ISBN:978-959-250-652-7.
• Ricardo Osés-Rodríguez, Meylin Otero-Martin, Nancy Ruiz-Cabrera, Rigoberto Fimia-Duarte, José Iannacone (2018) Prognosis for Hurricane Irma Through Regression Objective Regression and Its Impact on The Vector Populations at The Meteorological Station of Caibarién, Villa Clara, Cuba. Biotempo 15(1): 23-30.
• Osés-Rodríguez Ricardo, Fimia-Duarte Rigoberto, Osés-Llanes Claudia, Wilford-González Frank M (2022) Chaos Theory of Mathematics as seen from a New Perspective for Weather Forecasting. Bioscience Biotechnology Research Communications 15(3): 390-398.
• Rodriguez RO, Duarte RF, Meneses AG (2021) Modeling of the Number of Cold Fronts in Cuba Using the Objective Regressive Regression (ROR) Methodology; Impact of Sunspots. J Biomed Res Environ Sci 2(9): 870-875.
• Lorenzo Diéguez Fernández, Ricardo Oses Rodríguez, Rigoberto Fimia Duarte (2023) Impacto de las variables climáticas en los focos de Mosquitos en Camagüey, Cuba.” Centro Meteorologico Provincial de Villa Clara, Cuba.
• Jorge Luis Alonso Freire, Nibaldo Hernández Mesa, Ricardo Osés Rodríguez, Rigoberto Fimia Duarte, Lourdes María Basanta Marrero, et al. (2023) Efecto de la variación de la temperatura ambiental en la enfermedad cerebrovascular. Acta Medica del Centro 17(3).