Flooding
Student Name
Course
Date
Introduction
Flooding in rural areas is becoming a major hazard in many regions in terms of both frequency and magnitude. Floods adversely affect infrastructure, livelihoods, and production, slowing the accomplishment of sustainable development goals. The governments have established Early Warning Systems for torrents that appropriately address the system’s different components while being impact and community-based(Jamshed et al., 2019). Early warning systems are a critical pillar of disaster management as it is an integrated system of monitoring hazards, projecting, communication, disaster risk assessment, and preparedness and procedures that empower communities, individuals, businesses, governments, and other agencies in reducing the disaster risks in the progression of hazardous activities.
The systems are based on various aspects, including; the courses of flooding in Rural Areas, including heavy rainfall, which can contribute to multiple weather conditions. An example is tropical cyclones which can change into hurricanes and later cause heavy rain that can cause floods and even death. Besides that, overflowing rivers can also because floods in case of an overflow (Qiu et al., 2021). Properties near rivers can be protected by installing an Aqua-Barrier inflatable water dam that is cost-friendly and simple to manage. The other cause of flooding in rural areas is collapsed Dams. This happens if the barriers start to collapse and discharge a lot of water downstream, causing flooding (Bukvic&Harrald, 2019). This can be a major problem for people living in low-lying areas. This can destroy property as well as death. Also, snowmelt is another cause after melting snow or ice in colder climates. This causes many springtime streams to flow into rivers. Heavy precipitation has a high probability of flooding during the subsequent warm seasons. Snowpack starts to melt, releasing water that flows to the river, causing overflowing. Moreover, climate change also causes flooding, which is contributed heavily by uncontrollable human activities. Among the human activities include deforestation, increasing the carbon dioxide levels in the atmosphere that causes climate change causing flooding.
Risk knowledge
This aspect relates to the local assessment of the underlying risk of flooding, which involves analyzing hazard scenarios and vulnerability for evaluating risks to the elements involved with a detailed resolution. This is essential during the implementation of bottom-up and top-down strategies. Local risk assessment is normally addressed through engagement for the main settlements prone to flooding, including flooding areas and hazard threshold. Hazard mapping is another critical element for managing and responding to floods in rural areas. They offer insight into scenarios that need to be established to pinpoint exposure to the magnitude of different hazards. The hazard map provides maps of the flood hazard scenarios analyzed and a corresponding atlas of the areas prone to flooding. Vulnerability mapping, on the other hand, is a term that provides a detailed list of all vulnerable elements and core infrastructure that need to be updated periodically and documented. This aspect is responded to via the involvement of municipal government agencies in georeferencing, listing, and describing assets prone to destruction in the main localities.
Monitoring and Warning Components
The elements involved in accurate and timely forecasts ensured that the data collected was superior in quality and provided in real time. The second component is hydrological forecasts which guarantee a lead time of ten days. However, their accuracy tempts to be lower compared to hydraulic projections grounded on the capacity of the upstream hydrometer, which can take a lead time of up to twenty-eight hours. With the advancement in research, the accuracy of the hydrological forecast has been improved greatly by incorporating real data collected at the measuring stations (Jamshed et al., 2020). Another aspect is the impact-based threshold, which necessitates that warnings be prepared and dispatched on anticipated impact severity, enabling the end-users to implement suitable risk mitigation activities. Geographic-specific warnings offer a dense network performing extensive monitoring to guarantee that there is a positive coverage of the projection. This ensures that being more precise and dispatching warnings promptly will improve the ability to respond to disasters.
Aspects Regarding Dissemination and Communication
The different aspects used in distributing and communicating flooding hazards in rural areas include effective standard operating procedures. This tool is established by government policy, which takes the responsibility of developing the pathway that will be used in warning dissemination and the regulations for describing particular impact-based warnings (Longman et al., 2019). The main challenge faced by the disaster response team is effectively involving the different stakeholders and resolving conflicts of responsibility and mandate. The National Alert Code is undoubtedly developed, although various organizations claim to hold a similar mandate on flood risk administration. Over the years, dissemination and communication plans have been established in alignment with the National Alert Code and the shared outcome, which has been evidenced to offer similar information to stakeholders. This further leaves other dissemination to individual internal processes. As a result, the different stakeholders within the alert chain participate in precise communication channels. Another aspect is the timely and complete dissemination of flooding warnings in rural areas, which must ensure they accomplish the affected community members, including those residents in the most remote areas. The dissemination must be done promptly to prevent massive loss of lives and property.
The use of numerous dissemination channels is implemented to the exposed communities, which are warned through diverse media depending on their ability to use them effectively. The hazard response team erects a functional communication system to ensure that the emergency information will reach the relevant individuals within the right time and in numerous ways. The different levels involved include the municipal level, where the mayor has the freedom to implement diverse communication techniques, including voice calls and smartphone messages, to reach village leaders such as chiefs. Consequently, end-user dissemination and communicating agendas such as messaging content will reach the intended audience by tailoring it to suit the needs of the occupants of the rural area and ensure that they can understand how to interpret these laws.
Response Abilities to Communities Preparedness and Action
The availability of research on different aspects plays a great role in influencing impact and advice. This arises as messages comprise information providing insights on the anticipated impacts and providing direction on responding to and implementing risk mitigation initiatives. Other tools, such as roving seminars regarding the risk of flooding, are effective concerning different expectations. Another aspect is providing volunteers and community education. In this case, the volunteers’ goal is to guarantee coordination at the primary level, which involves helping the affected community members respond to notifications. The volunteers play the role of observing and performing coordination at the local level, educating the villagers on how to analyze alerts from the alert system and implementing swift actions to avoid mass destruction (Papilloud&Keiler, 2021). At the same time, contingency and preparedness plans are performed using the vulnerability and hazard map relied on by the emergency response team as a tool that can be utilized in benchmarking to improve response to flood disasters. Similarly, these tools are used to coordinate the process of responses to alerts. The state government is also strengthened by having access to flood maps at various scales, which is important in identifying the risk assets that are at risk in every village. The flooding maps also feed in-depth knowledge to the state government officials on the measures that can be implemented to reduce the impact of risk for the particular village. However, the state government’s challenge is maintaining an updated list of the different flood scenarios to identify their respective cause. Consequently, local community involvement creates a scenario where the end-users have a platform where they can actively make contributions to the different components of Early Warning Systems. Community members are key players in local risk assessment by providing information such as providing information on the fluctuating levels of vigilance, monitoring the possibility of the flood at a local level, and defining the relevant contingency actions to reduce the risk.
Sampling and Sampling Procedures
The sample for this study was limited to N = 160 counties located across the eight states (i.e., Louisiana, Mississippi, Arkansas, Alabama, Missouri, Tennessee, Kentucky, Illinois) of the Mississippi Delta Region.A binary logistic regression (two-tailed) power analysis using G*Power was needed for the study (Faul, Erdfelder, Lang, & Buchner, 2007).There are nine states was set as covariates as 1 to represent independent variables. Flooding and neighborhood disadvantage access, the odds ratio was set to 0.61, which was a small-to- moderate effect size which representing a .559/.441 probability level. O and 1, respectively represent the variables. The parameters show that based on the total sample size required to achieve the adequate statistical power was N = 160. Harrell, 2016;Hosmer, Lemeshow& Sturdivant, 2013 mentioned the general rule of thumb of 15 cases per one predictor is slightly higher.
Eight counties will be randomly selected from each of the eight-state located in the Mississippi Delta Region proportionally. Table 1 presents the number and percent of MDR counties by state and the associated number of counties per state that will be selected for the study. Counties by state were randomly selected by (a) assigning a number to each county; (b) setting on online random number generator to the total n MDR counties per state (i.e., for the state of Louisiana, setting the range of random numbers between 1 and 53); and (c) using the random number generator to select the required sample of counties per state.
Table 1
Proportional Random Selection of County Sample (N = 160)
| State |
N (%)
Counties in MDR |
N
Random County Sample |
| Louisiana | 53 (21%) | 34 |
| Mississippi | 47 (19%) | 30 |
| Arkansas | 42 (17%) | 27 |
| Missouri | 29 (11%) | 19 |
| Tennessee | 22 (9%) | 14 |
| Kentucky | 22 (9%) | 14 |
| Alabama | 20 (8%) | 12 |
| Illinois | 17 (6%) | 10 |
| Total | 252 | 160 |
Procedures for the Collection of Data
The study does not involve the recruitment of human subjects; as such, informed consent is not relevant to this study. County-level archival data from various sources, all of which are in the public domain and free to use for research purposes will be used. 2016 data from the U.S Census will be used as independent variables and the covariate of county population.The dependent variable of presence/absence of floodingwill be obtained from the CDC (2017) modified the Disadvantaged Flood Environmental Index, measured at the census tract and county level.
Instrumentation and Operationalization of Study Variables
Archival data was used from two sources:the U.S. Census and the CDC mFMD datasets. Data from the U.S. Census have historically and consistently been utilized in studies across a variety of disciplines (i.e., criminology, sociology, psychology, epidemiology, public health, urban studies). The U.S. Census Bureau 2016 setsdata collection standards and conducts evaluations to ensure that the data collected is valid, accessible, and meaningful for researchers and scholars (U.S. Census Bureau, 2016). 5% of the U.S. Census of the American population each year disseminated the most up-to-date and accurate estimates on census data (U.S. Census Bureau, 2016).
The CDC (2017) first developed the Flood MississippiIndex dataset in 2011, to increase empirical research on and subsequent awareness of geographical health disparities regarding access to flooding.The goal of the mFMDprogram was to “estimate access to awareness to flooding in the Mississippi Delta across the United States and regionally” (CDC, 2017. para. 5). CDC mapped the locations of flooded tract areas and county levels using a list of 54,666 flooded areas obtained from two national retail flood directories, InfoUSA and Mississippi Disaster Relief Program (CDC, 2017). The CDC used these data to calculate, for each census tract and county in the United States, an mFMDindex score, denoted as: (CDC, 2017). The CDC then published census tract and county-level mFMDdata mapped for each state.
Independent Variable 1: County-level poverty rate. The independent variable of county-level poverty rate was assessed using 2016 U.S. census data that assessthe county-level percentage of individuals living below the poverty level. This is a ratio variable. The possible range of scores was 0% to 100%, with a higher percentage indicating a higher county-level percentage of individuals living below the poverty level.
Independent Variable 2: County-level percentage of African American residents. The independent variable of county-level percentage of African American was assessed using 2016 U.S. census data that measure the county-level percentage of individuals classified as African American. This is a ratio variable. The possible range of scores iswas 0% to 100%, with a higher percentage indicating a higher county percentage of African American individuals.
Independent Variable 3: County-level percentage of elderly residents. Empirical literature has defined an elderly person as someone who is age 65 or older (Addington, 2013; Orimo, Ito, Suzuki, Araki, Hosoi, &Sawabe, 2006; Sabharwal, Wilson, Reilly, & Gupte, 2015). The independent variable of county-level percentage of elderly African American was assessed using 2016 U.S. census data onthe county-level percentage of individuals who are age 65 or older. This is a ratio variable. The possible range of scores was 0% to 100%, with a higher percentage indicating a higher county percentage of elderly persons, that is, persons age 65 or older.
Independent variable 4: County-level vehicle ownership. Another community risk factor denoted in the DFS conceptual model (Pothukuchi et al., 2008), the independent variable of county-level vehicle ownership will be measured using 2016 U.S. census data on the percent of occupied households that have one or more vehicles. This is a ratio variable. The possible range of scores is 0% to 100%, with a higher value indicating a higher percent of occupied households with one or more vehicles.
Independent variable 5: County-level crime rate. A DFS conceptual model (Pothukuchi et al., 2008) community risk factor, the independent variable of Type I (violent and property) crime rate will be measured using FBI UCR data for 2014. The offenses included under the Type I category are murder, rape, robbery, aggravated assault, burglary, larceny, auto theft, and arson (FBI, 2016). This is a ratio variable. The possible range of scores is 0% to 100%, with a higher value indicative of a higher Type I crime rate.
Dependent variable: County-level presence/absence of preparation for flooding.The dependent variable of county-level presence/absence of flood preparationwill be measured using CDC (2017) modified Flood Mississippi Data (mFMD) data. The mFMDindex is calculated by dividing the number by the number of people who were prepared for the flood and not preparedtimes 100 (CDC, 2016). These data are available at the census tract and county levels. As many of the counties in the Mississippi Delta Region lack preparation for flooding (CDC, 2016), this variable will be treated as dichotomous, coded where 0 = lack of flood preparation and 1 = presence of flood preparation.
This study focused on the relationships between Pothukuchi et al.’s (2008) five community risk factors and the flooding characteristic of presence/lack of flood preparation in a theoretically-valid location of the MDR. In this study, 160 counties in the MDR will be used as study participants. The use of proportional stratified random sampling of counties precludes the need to control for covariates. Temporal precedence will be addressed by using 2016 U.S. Census and 2016 CDC mFMDdata. Seven research questions are posed for this study.
RQ1. The 2016 county-level poverty rate significantly associated with the 2017 presence/ absence of preparation for floodingin 160 MDR counties?
Ho1. The 2016 county-level poverty rate significantly is not significantly associated with the 2017 presence/ absence of preparation for flooding in 160 MDR counties.
Ha1. The 2016 county-level poverty rate is significantly associated with the 2017 presence/ absence preparation for flooding in 160 MDR counties
RQ2. Is the 2016 county-level percent of African American residents significantly associated with the 2017 presence/absence of preparation for flooding in160 MDR counties?
Ho2. The 2016 percentage of African American residents is not significantly associated with the 2017 presence/absence of preparation for flooding in160 MDR counties.
Ha2. The 2016 percentage of African American residents is significantly associated with the 2017 presence/absence ofpreparation for flooding in 160 MDR counties
RQ3. Is the 2016 county-level percent of elderly (i.e., age 65 and older) householders significantly associated with the 2017 presence/absence of preparation for flooding in160 MDR counties?
Ho3. The 2016 county-level percent of elderly (i.e., age 65 and older) householders is not significantly associated with the 2015 presence/absence of preparation for flooding160 MDR counties.
Ha3. The 2017 county-level percent of elderly (i.e., age 65 and older) householders is significantly associated with the 2015 presence/absence of preparation for floodingin 160 MDR counties.
RQ4. Is the 2016 county-level vehicle-ownership rate significantly associated with the 2017 presence/absence of preparation for floodingin 160 MDR counties?
Ho4. The 2016 county-level vehicle-ownership rate is not significantly associated with the 2017 presence/absence of preparation for floodingin 160 MDR counties
Ha4. The 2016 county-level vehicle-ownership rate is significantly associated with the 2017 presence/absence of preparation for flooding in 160 MDR counties.
RQ5. Is the 2016 county-level crime rate significantly associated with the 2017 presence/absence of preparation for floodingin 160 MDR counties?
Ho5. The 2016 county-level crime rate is not significantly associated with the 2017 presence/absence of preparation for flooding in 160 MDR counties.
Ha5. The 2016 county-level crime rate is significantly associated with the 2017 presence/absence of preparation for floodingin 160 MDR counties.
All study data will be entered manually into an SPSS 25.0 software data file. Each randomly selected county will be entered as the study participant variables. A string variable will be created to record the associated state. Data from the U.S Census and mFMRdatasets that correspond with the county will be entered into the data file for the study independent and dependent variables. The data will be reviewed and adjusted for any entry error. As data will come from national data sets on county information, there will be no missing data.
Prior to conducting a binary logistic regression to test study hypotheses, preliminary statistics will be conducted. Descriptive statistics (i.e., mean, standard deviation, minimum and maximum scores) will be computed on all independent variables and the covariate of county population. Frequencies and percentages will be reported for the dependent variable of presence/absence of flooding preparation.
Binary logistic regression has few assumptions. The most concerning assumption is lack of multicollinearity between the independent variables (and covariate). Variance inflation factors (VIFs) will be computed to determine if multicollinearity is present; a VIF equal to or greater than 10.00 is indicative of multicollinearity. If any variables show multicollinearity, they will be removed from analyses. The overall significance of the binary logistic regression model will be determined by its model chi-square value, with significance set at p< .05. The Hosmer-Lemeshow chi-square will be included as an additional indicator of model significance. Unlike most statistics, a non-significant (i.e., p> .05) Hosmer-Lemeshow chi-square value indicates that the overall binary logistic regression model is a good fit to the data. Effect size will be documented by the NagelkerkeR2. Specific statistics will be reported to indicate if the individual predictor variables are significant. The variable’s Wald statistic and corresponding significance set at p< .05 will be reported. The odds ratio and 95% confidence interval of the odds ratio will be reported for each predictor variable as an indicator of effect size.
Survey
Please circle all that apply
- Are you prepared for flooding in your area? Yes or No
- Have you been affected by a flood in your area? Yes or No
- How long have you been in your home? Less than 5 years or More than 5 years
- Do you ______ your home?
- Own
- Rent
How long have you lived there? Less than 5 years or More than 5 years
- If your home was flooded, how do you estimate the damage? Partial or Full
If you have not experienced damage Skip to question 7
- Do you have flood insurance? Yes or No
- Have you gotten information for preparing for flooding at your local city hall? Yes or No
- Is your home in a flood zone? Yes or No
- Do you have flood insurance? Yes or No
- Do you feel that you are better prepared for flooding after experiencing flooding? Yes or No
