Options
Siti Aisyah Zakaria
Preferred name
Siti Aisyah Zakaria
Official Name
Siti Aisyah, Zakaria
Alternative Name
Zakaria, S.
Zakaria, S.A.
S.A., Zakaria
Main Affiliation
Scopus Author ID
56442195200
Researcher ID
EII-3879-2022
Now showing
1 - 3 of 3
-
PublicationSpatial max-stable with cyclic generalizes extreme value model for extreme ground level ozoneThe high level of ground-level ozone (GLO) concentration has serious adverse effects on health problems and affects the environment. This study integrates the spatial max-stable processes with the cyclic generalized extreme value (GEV) model to analyze and forecast extreme GLO levels. Spatial extreme provides a framework for analyzing and modelling the behaviour of rare events considering the extreme data pattern and the characteristics of several stations. One key component in this approach is selecting an appropriate GEV marginal distribution based on the data structure of each monitoring station. The choice of marginal distribution depends on whether there is an apparent trend in the extreme data series. In cases where the data exhibit strong seasonal variation, a stationary model may not be appropriate. This study acknowledges the seasonal variation in GLO data at various monitoring stations, influenced by the interchange of monsoons. Therefore, a seasonal variation model is considered as the marginal distribution for the spatial extreme model. Additionally, the study extends the non-stationary model from univariate cases to the spatial extreme model. This model incorporates a cyclic pattern in location parameters to complement the GEV distribution as a new marginal distribution within the max-stable process, a standard dependency model for spatial extreme observed at different locations. The clustering process using hierarchical cluster analysis (HCA) found that all the stations can be grouped into two clusters depending on the same characteristic of the weekly maxima data. The extremal coefficient between 1 to 2 indicates that the stations are dependent on each other’s within the cluster. Validation of the developed model is crucial for accurate predictions. Synthetic data approximating real data characteristics are generated to validate the model and facilitate predictions of future extreme cases based on return values for specific return periods. Return levels, indicating the average amount of extreme events within a specified return period, are used to predict GLO concentration levels across different locations, enhancing the understanding of GLO concentration patterns based on location categories. The presentation of return level results in return level mapping further aids in visualizing and interpreting the predictions for all monitoring stations. The main finding of this study indicates that the return level of GLO concentration increased as the return period increased. The results show that most of the return levels exceed the guideline of MAAQG for 8-hour average that is 0.06 ppm. Notably, the station in Kota Bahru (CA22) stands out with the lowest return levels, while the Shah Alam (CA25) station exhibits the highest estimated values. This high GLO concentration in Shah Alam may be attributed to its urban location, marked by high traffic density, industrial operations, and diverse meteorological influences. In conclusion, this study is highly significant as it offers valuable insights that can be applied in the fields of environmental and climatology, specifically regarding GLO in peninsular Malaysia. The methodology detailed in this study can be adapted for the analysis of other extreme datasets.
32 4 -
PublicationThe comparison of extreme rainfall prediction for Northern region of Peninsular Malaysia based on GEV and GPD models(AIP Publishing, 2024)
-
PublicationTeaching statistics with excel: a hands-on approach for engineering students to promote thinking skillsStatistics education has become increasingly important in today's data-driven world, as the ability to analyze and interpret data is critical in many disciplines. However, introductory statistics courses traditionally emphasize rote calculations and procedural knowledge, which can result in passive learning and disengagement from students who may not see the relevance of statistics to their engineering field. To address these challenges, this paper proposes using Excel worksheets as student learning materials in an introductory statistics course to shift from traditional to experiential learning. Excel worksheets provide a hands-on approach to learning that gives students the experience of the actual process of doing statistics. The Excel worksheet facilitates quick and accurate calculations, allows more time for students to interpret statistical results, and encourages active learning. The Excel worksheet allows for real-world data analysis and what-if analyses, making abstract concepts more accessible. In addition, the Excel worksheets are designed to promote 21st-century thinking and collaboration skills, which are increasingly important in today's workforce. This paper presents several examples of Excel worksheet designs for teaching descriptive statistics, developed using the framework of substitution, augmentation, modification, and redefinition (SAMR) model. Excel worksheets promote deep learning and facilitate students' understanding of statistical ideas, concepts, and methods through learning by doing. The paper concludes that Excel worksheets offer a valuable tool for teaching introductory statistics to engineering students, enhancing their thinking skills, and preparing them for the data-driven demands of their field.