AN ANALYSIS OF NOAA MONTHLY AND DAILY PRECIPITATION DATA TO IDENTIFY LONG-TERM TRENDS SINCE 1900 IN THE TENNESSEE RIVER BASIN
We chose stations within the Tennessee River basin defined by latitude 37°34’ - 34°04’ N and longitude 90°18’ - 81°12’ W. Stations that did not meet 100 years of data were removed from the dataset. In addition, seven stations were excluded due to large data gaps, leaving 121 stations that met these specifications.
We utilized the Mann-Kendall statistical test to detect increasing or decreasing trends in precipitation data. To apply this over the course of one year, each station’s records were first processed in a program to separate each year into 12 months. The Mann-Kendall test was then processed in an auxiliary program for the monthly datasets.
Results from the analysis show that monthly average precipitation trends (p < 0.05) increased at a greater frequency than they decreased. However, most stations showed no trend in monthly average precipitation. The months with the highest percentage of stations showing increasing trends were May (36%), October (28%), and November (27%). Decreasing trends were found for stations during the months of July (8%), February (3%), and August (2%). The majority of stations showed no significant trends whatsoever, ranging from 65% of stations in November to 98% of stations in March. During six months (Jan-Apr, Jun, and Dec) more than 90% of stations showed no trend in monthly average precipitation over the last 100 years.
The daily and hourly data will later be used to better understand high intensity storm events. Statistical analysis has yet to begin on that portion but it is the focus of future work.