October Weather Over the Past 100 Years in Douglas County KS
Is climate change evident in one of the most fluctuating months for weather?
Madeline Anderson, Natalie Hammer, Delaney Burns, Myjae Maloy
PSYC 500 Final Project
Embedded Files
Introduction
Explain the background of a real-world issue and problem your team aims to understand.
October is one of the months with the most variable changes in weather, especially in the Midwest. With Climate Change increasing over the last 200 years, our team wanted to analyze if global warming could be seen in October's temperature data in Douglas County over the last century.
Describe a couple examples of the issue.
Climate Change is an issue that is relative to more than just Douglas County, but the whole country as well. Fluctuating temperatures can negatively affect agricultural farms by resulting in drier soil, unpredictable water availability, and ultimately stopping crop fields from being properly irrigated. This can directly impact farms and specifically reduce overall yielding of crops in Douglas County such as corn, wheat, and more.
The economy is also at risk for a negative impact from Climate Change. Specifically in Douglas County, people depend on sectors and systems that are destined to be impacted by continuous Climate Change. With the anticipation of regional droughts, water supply shortages, and agricultural disruptions, crop production and state revenue and will inevitably decrease. This will result in higher prices for goods, reduced income, and job losses both statewide and nationwide.
This issue also has a large impact on air quality. Climate Change causes higher levels of greenhouse gases and pollutants in the air. This consists of harmful gases like carbon dioxide (CO2), methane (CH4), and particles suspended in the air such as mold spores, smoke particles, and atmospheric dust.
Briefly address previous studies that investigated the issue and limitations of the previous studies.
The Kansas Health Institute investigated a high likelihood that Climate Change can have a severe impact on human health. This study shows how changes in temperatures majorly affect air quality, there will be a longer pollen production period that will expand allergy season and worsens asthma symptoms, and overall a longer season with mosquitos and other insects capable of transmitting diseases will only put human health at risk. This study projects that eastern Kansas specifically will be impacted by Climate Change with continuous warmer temperatures. This issue was further found to increase rates of cardiovascular diseases, respiratory diseases, and other chronic health conditions. A limitation of this study is that these health risks are formulated by a collection of various methods of research from a restricted area. Although the data collected is credible, this study is based on weather patterns and climate changes only in Kansas. These precise quantitative projections are not reliably applicable to every state or region in the country. (Corbett)
The United States Environmental Protection Agency analyzed how impactful Climate Change can be in the state of Kansas. This research shows that the soil is becoming drier, and changes in the timing and size of rain precipitation is altering crop yields. Approximately 22% of the farmland in Kansas is irrigated, and with soils continuing to get drier, the need for farmers to irrigate their crops will only increase. This is an issue for farms because sufficient water is steadily declining in availability. Specifically, this study shows that the amount of water stored in these irrigation systems has declined by more than 25% since the 1950s. Climate Change has a severe impact on agriculture and crop production.
A limitation from this study is that the data collection is mainly focused on how precipitation and water sources are affected by Climate Change. This study does not have as big of an influence on change of temperatures over time. There are many factors that cause Climate Change and are affected by it, and this study is limited by focusing on one factor. ("United States Environmental Protection Agency")
Data Curation and Ethics
Describe the source of data and why your team curated the particular dataset(s).
The data used for these analyses were curated from the National Oceanic and Atmospheric Administration Climate at a Glance County data page. We downloaded data sets of Average Temperature, Maximum Temperature, Minimum Temperature, and Precipitation in Douglas County from October 1920-2020. All temperature measurements were provided in Fahrenheit and precipitation was measured in inches. We chose these data sets because we wanted a sufficient amount of different variables to complete our analyses on.
Discuss ethics that entail obtaining the data.
We chose this data because we wanted to get our data from a trusted website and Climate at a Glance is a government program. We did this because we don’t want to collect false data or mislead people with the data we have. It would be unethical to present results that are not true or supported. The information obtained in this data set mostly consist of temperature and dates, plus the precipitation data. There should not be any ethical issues because this dataset is public and does not include any personal information that would need to be protected by anonymity, informed consent, or other ethical protections.
Data Preparation and Exploration
Describe the variables in the data frame(s) and the data type of each variable.
The average temperature variable is the average of the temperatures recorded for each day that year in the month of October. The minimum temperature variable is the lowest daily temperature recorded for that year in the month of October. The maximum temperature variable is the highest daily temperature recorded for that year in the month of October. The precipitation variable was the sum of all precipitation recorded in the month of October for that year. All four of these variables are stored as floats. The year and month data, with month always being 10, are integers.
Describe the size and shape of the data frame(s).
Before combining the data, each data set had 101 rows and 2 columns, one containing the month and year, and the other containing the values for that variable. After we combined the data into one data frame and added a column for the temperature range, there were 101 rows and 7 columns.
Reshape or combine data frames if necessary.
The combined data frame's 7 columns are Average Temperature, Maximum Temperature, Minimum Temperature, Precipitation, Year, Month, and Range, which is the Maximum Temperature minus the Minimum Temperature.
Post data visualization and describe insights from data visualization.
Our Average Temperature data appears to be normally distributed and has a mean of 57.41 and a standard deviation of 3.23. Our Maximum Temperature data appears to be normally distributed and has a mean of 69.13 and a standard deviation of 4.12. Our Minimum Temperature data appears to be normally distributed, with a bit of a skew. The mean is 45.68 and the standard deviation is 2.92. Our Precipitation data appears to be mostly normally distributed, but clearly skewed. The mean is 2.97 and the standard deviation is 1.95. Our Range data appears to be normally distributed, but has a flatter curve. The mean is 23.46 and the standard deviation is 3.06.
Model Building and Validation
Report summary and descriptive statistics of each variable - means, SDs, counts, etc.
Report the results of an analysis that checks the normality of the probability distribution of at least one continuous variable.
Average Temperature appears to be normally distributed with some variation. This makes sense because this data is the average of a naturally occurring phenomenon, temperature. Precipitation has a bit more variation than most of our other variables, though it follows the normal CDF close enough to be considered mostly normal. Because Range is a calculation between the Minimum and Maximum variables, this variable should be more normally distributed than each was. We believe that is the case here.
Model Building and Validation
Report the procedure and results of a linear regression analysis with two continuous variables.
We completed a linear regression analysis between the two continuous variables of Average Temperature and Year. Our linear regression showed a best-fit line with a slope of -0.019. We then conducted a pairs bootstrap to determine the significance of this slope. We chose to generate 10,000 bootstrap replicates in this analysis. It was determined that our slope was not significant, as our 95% confidence interval included 0 ([-0.041, 0.003]). Therefore, there was no significant relationship between Average Temperature in October in Douglas County and Year.
Report the procedure and results of permutation/bootstrap hypothesis testing with a discrete and continuous variable, if appropriate.
A permutation hypothesis test was then done to address the question of whether there is a significant change in average temperatures before and after a certain point. For us, that certain point was 1980, where we split our data into the 40 years before (and including) and the 40 years after. We chose 1980 as the year to split our data based on information from the Clinton White House and the NOAA Climate.gov website. Both sources provided proof that the warmest years and highest record-setting temperatures have all occurred since 1980 ("Climate Change", Lindsey). Two tests were done, one with the difference of means as the test statistic and the other with the difference of standard deviations as the test statistic. The difference of means test had a p-value of 0.025 that was significant (alpha=0.05). Therefore, the mean temperature in October has been significantly higher after 1980 than before 1980. However, the difference of standard deviations test did not have a significant p-value.
Report any additional analysis your team conducted.
We also conducted a permutation hypothesis test for the range of temperatures before and after 1980. We completed two tests here as well, with one using difference of means as the test statistic and the other using difference of standard deviations. Because this difference of means test had a p-value of 0.029 that was significant (alpha=0.05), we can conclude that the range of temperatures observed in October before 1980 is larger than the observed range of temperatures after 1980. Again, the difference of standard deviations test did not have a significant p-value.
Discussion
Briefly summarize the objective of the project, data curation and preparation, exploratory data analysis, and model building and validation in one short paragraph
The objective of this project is to see how the average temperature and precipitation in Douglas County for October has changed over the past century. We obtained our data from NOAA Climate at a Glance and prepared the files in Google Colab by determining the type, shape, and size of the average temperature, the minimum temperature, the maximum temperature, and of precipitation. We compared the averages of both temperature and precipitation in Douglas County for the month of October from 1920 to 2020. We curated the data by combining the the min, max, and mean of temperature and precipitation along with the normality of precipitation and the normality of the maximum, minimum, and average temperature for the years 1920-2020. We performed five linear regression analyses. The first compared the maximum temp and the minimum temperature, the second compared average temperature and the year, the third compared precipitation and the year, the fourth compared average temperature and precipitation, and the last compared the range of temperature and the year. We also performed a bootstrap sample that compared the average temperature and year. Furthermore, we permutated the average temperature means, the average temperature standard deviations, the temperature range means, and the temperature range standard deviations for the years before 1980 and after 1980.
Describe informed insights
Our results for the paired bootstrap sample showed that there was no statistically significant difference between average temperature and year. However, the average temperature and year permutation difference of means test results showed that there is a significant difference in the average temperatures in October before 1980 versus after 1980. The average temperature and year permutation difference of standard deviations test results concluded that there is no difference between average temperatures in October before 1980 and after 1980. The range of temperatures and year permutation difference of means test results determined that there is a significant difference in the range of temperatures in October before 1980 versus after 1980. Lastly, the range of temperatures and year permutation difference of standard deviations test results showed no statistically significant difference in the range of temperatures in October before 1980 versus after 1980.
Our results show statistically significant differences between the mean temperatures before and after 1980 and the range of temperatures before and after 1980. For the mean temperature and year permutation difference of means test, the p-value could be considered significant at the alpha = .05 level, which is considered slightly significant compared to an alpha of 0.01 or 0.001. The same applies to the range of temperatures and year permutation difference of means test, which had a p-value of 0.029. The results of the first permutation test, that the mean temperature in October has been significantly higher after 1980 than before 1980, make sense, as climate change has caused an increase in temperatures across the globe. However, the finding that the range of temperatures has actually been smaller since 1980 is surprising, as the expectation due to climate change is that temperatures will be more fluctuating.
Discuss limitations and future directions
Given temperature changes throughout the day, the time that the temperature was recorded on each day could affect the reported temperature of that day and therefore affect the overall max, min, and mean values for temperature in October for Douglas County. There are also instrumental errors that play a part in the data we obtained. For some temperature measurements, the thermometer could have not been recording the accurate temperature, resulting in inaccurate measures of temperature. In the future, it would be interesting to see how temperature and precipitation in other parts of the United States has changed for the month of October over the past century. For example, take a sample of temperatures for the states in the South and compare them to a sample of temperatures for the states in the North, Midwest, East coast, and West coast.
Discuss implications of the results in terms of policy and community interventions
It is hoped that our results will raise awareness of how the climate is changing by giving an example that is so closely connected to KU students by using measurements from Douglas County. Some implications of the results could include passing bills that protect the environment and combat climate change.
Citations
“Climate Change Over the Past 100 Years.” Edited by Pres. Bill Clinton, National Archives and Records Administration, National Archives and Records Administration, 2000, clintonwhitehouse4.archives.gov/Initiatives/Climate/last100.html.
Corbett, S., Ph.D., Orr, J. M., M.P.H., & Hunt, C., M.P.H. (2019, July 22). Health and Climate Change in Kansas (July 2019). Retrieved December 06, 2020, from https://www.khi.org/policy/article/19-35.
Lindsey, Rebecca, and LuAnn Dahlman. “Climate Change: Global Temperature: NOAA Climate.gov.” Climate Change: Global Temperature | NOAA Climate.gov, National Oceanic and Atmospheric Administration, 14 Aug. 2020, www.climate.gov/news-features/understanding-climate/climate-change-global-temperature.
NOAA National Centers for Environmental information, Climate at a Glance: County Time Series, published November 2020, https://www.ncdc.noaa.gov/cag/.
United States Environmental Protection Agency. (2016, August). What Climate Change Means for Kansas. Retrieved December 06, 2020, from https://www.kansasforests.org/resources/resources_docs/climate-change-ks%202016.pdf.
Page updated
Google Sites
Report abuse