Air Pollution Analysis in Kuwait Using a Statistical Technique (CUSUM)

Investigating the changes in the air pollutants trends of an area is important as it helps in making further action plans for further implementation of control strategies. Time series analysis provides indication to analyze any effect of uncontrolled changes in pollutants. In this study, exponentially weighted moving average (EWMA) and cumulative sum (CUSUM) analyzing methods are applied for detecting the trends and change in air pollutant concentrations in Kuwait. CUSUM method is effective in detecting shifts from average mean obtained by EWMA technique. The study aimed to investigate trends in major pollutants in three selective areas in Kuwait during the past five years. The data obtained from three monitoring stations in the study areas Ali Subah Al-Salem, Al-Mutla, and Al-Mansouriya for carbon monoxide (CO), nitrogen dioxide (NO2), sulfur dioxide (SO2), particulate matter—less than 10 micrometers (PM10), hydrogen sulphide (H2S), and non-methane hydrocarbon (NMHC). Increase in CO and NMHC concentrations in the three areas and decrease in PM10, SO2, and NO2 concentration levels in non-industrial areas Al-Mutla and Al-Mansouriya are observed using CUSUM method.

The burning of fossil fuels and the decline of the world's forest areas have both directly affected a steady rise in CO 2 concentrations in the last half century. On the other hand, still there is no clear and uncertain impact on the long-term [5]. Britain's Meteorological Office and the USA's NASA Goddard Center for Space Research both confirm recent rises in global temperatures, but it is unclear whether these are due to levels of greenhouse gases or natural variations in global climate [6].
Kuwait's development policy plan aims at diversifying sources of income by encouraging the expansion of the economy and reducing the country dependency on oil exports. As a result of that plan, the industrial sector receives special support and encouragement from the government. As an example, the Shuaiba Industrial Area is a governmental authority located 50 km south of Kuwait city between Ahmadi south pier and Mina Abdulla along the costline, with an independent budget and attached to the Minister of Commerce and Industry. Shuaiba Industrial plants locate near the expanding residential projects such as Ali Subah Al-Salem area, Subah Al-Ahmad area, Jaber Al-Ahmed etc. Emissions from the various industries have increased continuously in Shuaiba over the last three decades [5]. The resulting impact on both the performance of the industries and the environment around Shuaiba is a cause for increasing concern to the State Authorities of Kuwait, the Shuaiba Area Authority and even local industry. This has led to a recognition of the need for the scientific community to develop a sound approach for assessing the wide range of health and environmental effects that result from exposure to toxic chemicals [7]. [8] proved the effectiveness of both techniques Cumulative Sum Control Charting and Exponentially Weighted Moving Average Control Charting. The study investigated that Cumulative Sum Charting provided slightly earlier alarms, and Exponentially Weighted Moving Averages are easier to use. In addition, the study noted that use of these techniques could allow detection of changes in time to mitigate the negative effects of the change and could be applied to a very wide range of processes. The purpose of this study is to investigate the ability of the Cumulative sum (CUSUM) technique to identify a step change in pollution levels in three areas in Kuwait. If appropriate, the technique could then be developed further for application in more parameters and areas in Kuwait. The exponentially weighted moving average (EWMA) and cumulative sum (CUSUM) techniques are applied to air quality data at three areas in Kuwait Ali Subah Al-Salem, Al-Mutla, and Al-Mansouriya. The site is characterized as monitoring stations in each area with daily concentration measurements. In Ali Subah Al-Salem, industrial plants are located nearby the area, while Al-Mansouriya area are surrounded by heavy traffic because of main rings and roads in Kuwait City. However, Al-Mutla is plain area with no major external exposure. The continuous monitoring is carried out by Kuwait Environment Public Authority, Kuwait. The 24 hourly data observed during 2013-2017 (study time) for CO, NO 2 , SO 2 , PM10, H 2 S, and NMHC is considered to analyse for any changes in the behavior of these pollutants. Environmental studies are essential before any plans especially the residential ones. This research investigates the trends of pollutants in three areas in Kuwait using CUSUM technique, which can be easily applied in several Environmental projects for several purposes especially in the design of new residential areas like Al-Mutla or other future development plans in Kuwait. In addition, the study investigates the uncontrol pollutants levels in order to apply some control program to decrease the bad effect of rising pollutants in some study areas.

Methodology
The CUSUM technique was developed initially for statistical quality control [9] [10] [11] [12]. CUSUM methods apply to observations recorded over time (daily, weekly, monthly). The observations may be physical measurements, counts or rates and may be grouped (in production batches for example) or individual observations (e.g., as here, daily average concentrations of a pollutant at a monitoring station).
In order to apply the CUSUM approach to air pollutants in Kuwait areas, the procedure is described briefly. First, we will consider each concentration meas- Moreover, the change in terms of increased or decreased process mean can be detected, respectively by computing the quantities as (10); where parameter k is the reference value to be chosen appropriately. For the normally distributed variables with mean 0 and unit standard deviation, slackness factor k can be chosen as 0.5 to indicate the shift of 1σ in process mean. The confidence limits specified for the CUSUM control charts are ±hσx, where h = 4 or 5 and σx is the standard deviation, slackness factor k can be chosen as 0.5 to indicate the shift of 1r in process mean. The confidence limits specified for the CUSUM control charts are ±hσx, where h = 4 or 5 and σx is the standard deviation [13].

Results
The study will be based on data from three areas in Kuwait from different environment conditions and different Governorates. There will be brief description of each of area in the case study. Capital Governorate (Al-Mansouriya) Mansouriya is an area in the Capital Governorate and a suburb of Kuwait City as shown in (Figure 2), its Population reaches 8352 in 2008 [2]. The total area is 1,255,302.329 m 2 . The area selected faces severe air pollution problem mainly from transportation and daily traffic emissions from huge number of vehicles passing by its surrounding roads. The study will evaluate the pollutants levels in this area from Environment Public Authority last 5 years for urban air pollution (PM10, CO, H 2 S, NO 2 , SO 2 , NMHC).
Al-Jahra Government (Al-Mutla) Al-Mutla is a series of plateaus located north-east of Kuwait, Al-Mutla area is 40 km away from the Kuwait capital.
About Future AL Mutla Housing Project: In order to meet the increasing demand for housing care for citizens, the Public Authority for Housing Welfare (PAHW) is undertaking a series of projects in the form of new large urban areas, outside the current Metropolitan Area is one such project.  Ali Subah Al-Salem area is located the South Kuwait ( Figure 2). The area is considered from the newest urban areas in the last years in that region. The area consists of 9 blocks with total population of 47,302.
A residential area of Umm al-Hayman (now known as Ali Sabah al-Salem) was planned, without taking into account the environmental and health effects associated with the site, which was chosen for the residence of hundreds of

Results and Discussion
The CUSUM and EWMA techniques are applied to air quality data at three areas

Trend analysis Results
The Minitab results for each pollutants trends are presented in this chapter.
The results will be investigated for each pollutant exceeding the average for each area in the study by calculating the percentage difference either percentage increase or percentage decrease. In addition, the major change in concentration levels are described in the tables.
The CUSUM results can be summarized in the following Tables 1-6 for each pollutants.

Conclusions
The EWMA method provides first alarm for uncontrolled behavior of pollutants   The CUSUM method is essential to identify the shifts from the mean of any process. CUSUM method is applied to detect the changes in air pollutant con-