Analysis of Kano Meteorological Data using Time Series Analysis and Empirical Orthogonal Functions
No Thumbnail Available
Date
2020
Journal Title
Journal ISSN
Volume Title
Publisher
FACULTY OF SCIENCE, KANO UNIVERSITY OF SCIENCE AND TECHNOLOGY, WUDIL
Abstract
In this work, two different statistical techniques were used to analyses the meteorological
data of Kano State. The meteorological parameters utilized are daily Solar Radiations,
Sunshine Hours, wind speed, Maximum and Minimum Temperatures, Rainfall, Cloud Cover,
and Relative Humidity, data spanning for thirty-one years (1980 to 2010). The statistical
analysis involves the use of time series analysis and empirical orthogonal functions. In the
time series analysis (TS) all the parameters are assumed independent. The second is the
Empirical Orthogonal Transformation (EOF) in which the data were analyzed using unrotated
and orthogonal transformations and six components were extracted. From component
matrices, it is discovered that there are two distinct season as Rainy and Dry seasons. The
rainy season has two components comprising heavy rain and light rain, while the dry season
comprises of three different types of seasons. The period of heavy rain is around 3.3 months
and period of light rain is 1.6 month. These gives a total of 4.9 months for rainy season and
7.1 months for dry season. The values of the eigen values are consistent with what is
observed in real-time of seven (7.1) months of dry season and (4.7) months of rainy season.
Now combining the two results, it can be said that the TS analysis show that all the
parameters have seasonal variation, and the EOF now described the two seasons
quantitatively
Description
Keywords
meteorological data, regression analysis, time series analysis, empirical orthogonal transformations, oblique transformations