Background Information

The daily global solar exposure is nothing but the total solar energy for a day falling on a horizontal surface. The monthly mean daily global solar exposure is the average of all available daily solar exposure in the month. The daily solar exposure is measured from midnight to midnight. The variability is observed in the monthly solar exposures across all over the world. There is a variation due to climatic changes. Typical values of the daily global solar exposure are ranges from 1 to 35 MJ/m^2 (mega joules per square meter). The values of the daily global solar exposure are highest during the cloudy conditions and it lowest during the clear sky days. The diffuse exposure is always less than or equal to the global exposures for the same period. Diffuse solar exposure is the total amount of solar energy falling on a horizontal surface from all parts of the sky apart from the direct sun. The solar energy received at the surface of earth can be separated in two basic components such as direct solar energy and diffuse solar energy. Diffuse solar energy is the energy receiving from sun beam. The beam of sun is quite intense, and hence they are describing as shadow producing radiation. The study of the global solar exposure is important due to its variation pattern based on the climatic conditions. We know that there are different climatic conditions exists throughout the year. In some months, there is more sunlight and sometimes there is a very low sunlight due to the cloudy conditions. If the sky is clear, then sunlight will be more. There are continuous changes in the solar exposures throughout the year. Also, there are some periodic variations observed in the solar exposures. It is observed that the solar exposure decreases from January to June and again it increases from June to December. It is observed that the mean monthly solar exposure decreases from the first month January to sixth month January and then again it increases from sixth month to twelve month December. Again, this cycle is continuous for next year. This cycle is same for all years. So, there is a cyclic movement of the mean monthly solar exposure data. The estimates of the solar estimates are useful in a wide range of the applications, mainly in the field of agriculture, engineering and research. Some examples for use of solar exposure estimates includes the monitoring plant growth and the disease control, evaporation and irrigation, architecture and building design, power station condenser cooling system, power generation, calculation of water requirements for crops, solar heating system design and use, skin cancer research, research into coral growth, weather and climate prediction models, solar powered car races, etc. The data regarding the solar radiation can be potentially provided in a variety of forms for suit of the given applications.

Reporting/Dashboards

It is important to study the variation patterns and trends in the data for the mean monthly solar exposure. By using this study we can understand the basic patterns of mean monthly solar exposure and we can use different estimates based on this data for the future prediction purpose. We have to check the variability pattern and trends as per months. For this research study, the research questions are summarised as below:

Whether the given data for mean monthly solar exposure shows any particular type of trend involved in the data? Is there any seasonal variation observed in the data for mean monthly solar exposure? Whether there are any periodic changes observed in the mean monthly solar exposure?

The proper methods of the data collection should be used for getting unbiased results. Use of random sampling techniques is useful if there is large population. But here, we do not need to use any type of sampling technique because we have to collect the entire data for the years 2001 to 2017. This is complete enumeration. The mean monthly solar exposure is nothing but the average of all available daily solar exposure for the month. The data for this study is collected from the Ballarat Hopetown Rd station number 89111. Data for mean monthly solar exposure is collected for the years from 2001 to 2017. All types of errors and mistakes are avoided during the process of data collection. The observations of daily solar exposure are collected from the midnight to midnight. All types of errors and mistakes are reduced during the process of data collection. If data contain the mistakes, there is a possibility of getting biased results.

For this research study, we have to use the different methods of analysis or research study. First step in every research study is to establish the research hypothesis. For this research study, the research hypothesis is stated regarding the variability in the data for mean monthly solar exposure. We want to check or test whether there is any variability or seasonal trends observed in the data for mean monthly solar exposure. For checking this we have to use the time series pattern and we have to use the graphical analysis for checking the variation. For this research study, the data is collected from Ballarat Hopetown Rd station number 89111. Data is collected for the year 2001 to 2017. Then by using the statistical tools and techniques, data is analysed for getting the results. The IBM Watson is used for the analysis purpose. We have to check the different types of variations involved in the data for solar exposure for the monthly data. We will check whether there is any seasonal variation exists in the data for solar exposure or not. We will check this seasonal data by using graphical method. We will use the time series plot for the study of variation of the mean monthly solar exposure. Apart from that, we will use some more analytics skills with the help of IBM Watson. We will use IBM Watson for producing different types of bar diagrams, time series plots, etc. We will analyse the different aspects of the data for the mean monthly solar exposure from the year 2001 to 2017. Data for year 2017 is not complete and data up to available months is collected for this research study.

Research

As discussed in the research methodology we have to use the descriptive statistics, graphical analysis and different tools and techniques of the time series analysis for the given Monthly solar exposure. We know that the descriptive statistics is very useful for understanding the nature and shape of the data under study. The mean of the data and standard deviations of the data explains the measures of central and measures of variation. Although use of these measures is not very useful for the analysis of time series plots, but it provides some initial idea about the nature of data. Also, we have to use the some graphical analysis for this research study. We have to use the time series plot, bar diagrams, etc for this research study. By observing time series plot, we get the judgement about the variation pattern and seasonal variation or periodic variation in the data set. Here, we have to see the variation pattern in the data set of the mean monthly solar exposure. First of all we have to see the descriptive statistics for the entire collected data for the Monthly solar exposure. From this table, it is observed that mean monthly solar exposure is given as 15.9949 with the standard deviation of 7.26174. The minimum monthly solar exposure is observed as 5.30 while the maximum solar exposure is observed as 30.80. Now, we have to see what the trend of the monthly solar exposure is over the given years.

From the above trend line it is observed that the monthly solar exposure shows continuous up and down movement for the given years. There are no high amplitudes present in the above diagram for the given years. If we observed the given data yearly, then we could not find so many variations in the data sets. If the data of the mean monthly solar exposures is used, then we can clearly show the variation patterns exist between the data. We have to see this variation pattern for the monthly solar exposure data.

Now, we have to see some discoveries regarding the data for the monthly solar exposure by using the IBM Watson. The screen shots of the discoveries are pasted below:

It is observed that there is a decrease in the mean monthly solar exposure from the year 2001 to 2017.  The predictive strength of the monthly solar exposure is given as 72%.

Recommendations for CEO

The prediction power of the prediction of monthly solar exposure by using monthly data is about 72% approximately.

The above graph shows the periodic variation exists between the mean monthly solar exposures by the year.

It is observed that the mean monthly solar exposure decreases from the first month January to sixth month January and then again it increases from sixth month to twelve month December. Again, this cycle is continuous for next year. This cycle is same for all years. So, there is a cyclic movement of the mean monthly solar exposure data.

The predictive strength for the monthly solar exposure is given as 38%. This means predictive strength is less and results would not be strong for this data.

It is observed that the monthly solar exposure from year 2001 to 2017 is increased by 1.2% only. This means it shows the steady and constant nature. There is no any significant difference in the mean monthly solar exposure for the first years and last years.

The above graph shows that there is decrease in the monthly solar exposure by year.

Now, we have to see the bar diagram for the monthly solar exposure data which is given as below:

The above bar diagram shows that there is no any specific pattern of distribution followed up by the monthly solar exposure data.

The trend of monthly solar exposure for the months is given as below:

From this diagram, it is observed that there is up and down periodic nature of the monthly solar exposure. The periodic trend is observed for the given data for monthly solar exposure.  From the above time series plot, it is observed that the mean monthly solar exposure decreases from the first month January to sixth month January and then again it increases from sixth month to twelve month December. Again, this cycle is continuous for next year. This means we conclude that there is a cyclic movement of the mean monthly solar exposure. We cannot use the linear relationship for the prediction of the further mean monthly solar exposure. We need to use different model for prediction of mean monthly solar exposure.

Conclusions

1. It is observed that mean monthly solar exposure is given as 15.9949 with the standard deviation of 7.26174. The minimum monthly solar exposure is observed as 5.30 while the maximum solar exposure is observed as 30.80.
2. It is observed that the monthly solar exposure shows continuous up and down movement for the given years.
3. It is observed that there is a decrease in the mean monthly solar exposure from the year 2001 to 2017.  The predictive strength of the monthly solar exposure is given as 72%.
4. It is observed that the monthly solar exposure from year 2001 to 2017 is increased by 1.2% only. This means it shows the steady and constant nature. There is no any significant difference in the mean monthly solar exposure.
5. The bar diagram shows that there is no any specific pattern of distribution followed up by the monthly solar exposure data.
6. It is observed that the mean monthly solar exposure decreases from the first month January to sixth month January and then again it increases from sixth month to twelve month December. Again, this cycle is continuous for next year. This cycle is same for all years. So, there is a cyclic movement of the mean monthly solar exposure data.
7. The prediction power of the prediction of monthly solar exposure by using monthly data is about 72% approximately.

References

https://www.bom.gov.au/jsp/ncc/cdio/weatherData/av?p_nccObsCode=203&p_display_type=dataFile&p_startYear=&p_c=&p_stn_num=89111

Crowder, R. B., 1995: Wonders of the Weather. Australian Government Publishing Service Canberra.

Gautier, C., Diak, G., Masse, S., 1980: A Simple Model to Estimate the Incident Solar Radiation at the Surface from GOES Satellite Data. J. of Appl. Meter. 19 1005-1012.

Weymouth, G., Le Marshall, J., 1994: An Operational System to Estimate Insolation over the Australian Region. Proc. Pacific Ocean Remote Sensing Conference 443-449.

Weymouth, G., Le Marshall, J. F., 2001, Estimation of daily surface solar exposure using GMS-5 stretched-VISSR observations: The system and basic results, Australian Meteorological Magazine, 50, 263-278.

Antony, J. (2003). Design of Experiments for Engineers and Scientists. Butterworth Limited.

Babbie, E. R. (2009). The Practice of Social Research. Wadsworth.

Beran, R. (2000). React scatterplot smoothers: Superefficiency through basis economy. Journal of the American Statistical Association.

Bickel, P. J. and Doksum, K. A. (2000). Mathematical Statistics: Basic Ideas and Selected Topics, Vol I. Prentice Hall.

Casella, G. and Berger, R. L. (2002). Statistical Inference. Duxbury Press.

Cox, D. R. and Hinkley, D. V. (2000). Theoretical Statistics. Chapman and Hall Ltd.

Degroot, M. and Schervish, M. (2002). Probability and Statistics. Addison – Wesley.

Dobson, A. J. (2001). An introduction to generalized linear models. Chapman and Hall Ltd.

Evans, M. (2004). Probability and Statistics: The Science of Uncertainty. Freeman and Company.

Hastle, T., Tibshirani, R. and Friedman, J. H. (2001). The elements of statistical learning: data mining, inference, and prediction: with 200 full-color illustrations. Springer – Verlag Inc.

Hogg, R., Craig, A., and McKean, J. (2004).  An Introduction to Mathematical Statistics. Prentice Hall.

Liese, F. and Miescke, K. (2008). Statistical Decision Theory: Estimation, Testing, and Selection. Springer.

Pearl, J. (2000). Casuality: models, reasoning, and inference. Cambridge University Press.

Ross, S. (2014). Introduction to Probability and Statistics for Engineers and Scientists. London: Academic Press.

Todd, G. (2007). Descriptive Statistics. Topics in Biostatistics. New York: Springer.

Huck, S.W. (2007) Reading statistics and research. 5th ed., Allyn & Bacon: Boston, MA

Pallant, J. (2007) SPSS survival manual, 3rd ed., Allen & Unwin.

De Vaus, D. (2002) Analyzing social science data. SAGE Publications: London.

Wild, C.J., & Seber, G.A.F. (2000) Chance encounters: a first course in data analysis and

inference. John Wiley & Sons: New York