Measuring, Modeling, and Understanding Orographic Precipitation Regimes and Hydrology in Mid-Latitude Mountain Regions

"PIs have set out an array of rain gauges along the Cataloochee Divide, Mt. Sterling, and the eastern part of the AT. During the summer of 2011, ten stations in the park were upgraded to include temperature and relative humidity sensors, to give a more detailed in situ view of the environment on the ground during these significant events. The duration and occurrence of rain events is greatly dependent of the rain gauge location and the daily record for each rain gauge illustrates the variability of mountainous area precipitation. During a 10-day period in July of 2008, rain gauges along the Cataloochee Divide recorded an almost 100% difference between different sites in the amount of rainfall, ranging from 2.5cm to 6.66cm. During 2009, one park gauge collected 3x as much rain in one event as another less than 8km away. Rainfall accumulation increases in the NE direction consistent with the track of Southwesterly events as they interact with the mountain ridges. In 2011, one convective event observed by the network occurred on July 14-15, when a stationary thunderstorm observed by two RGs in our network recorded almost 5"" of precipitation over four hours. This event caused significant damage to the Cherokee Tribal Fish Hatchery and to the Gunter Fork trail in the park. This was studied in more detail in study #GRSM-01085. Additional work improving the ability to forecast flash floods in these mountains is reported in IAR 2011. NASA's TRMM precipitation radar missed about 50% of the rainfall events detected by gauges and amount of rainfall for given events may have been off by plus or minus about 50%. One explanation for this is that raindrops greatly increase in size when passing through fog or other low clouds, leading to much higher precipitation levels on the ground than upper atmosphere observations predict. Data from the rain gauges can be used to change the parameters for the interpretation of the TRMM data, resulting in much higher accuracy. NEXRAD radar also did not match well with measured precipitation, especially at low precipitation rates. Preliminary results from paired high elevation and low elevation micro radar units showed evidence of the elevational (orographic enhancement) effect with a cumulated precipitation about 4-10 times higher on the mountain ridge (Purchase Knob) than in the valley nearby (Clyde). In addition, data indicate higher rain rates and longer duration for the rain events on the mountain ridge than in the valley. For the same storm system, there is large spatial variability of rain drop size distributions between ridges and valleys, and between exposed upwind ridges and the inner region. Summer precipitation is characterized by large event-to-event variability including a combination of stratiform (frontal systems) and convective (local thunderstorm) properties. During fall, stratiform precipitation dominates and rainfall is 2 times more frequent at the ridge than in the valley, corresponding to a 100% increase in cumulative rainfall at high elevation. Overall, rain events of less than 3mm per hour exhibit very little seasonal or yearly variability and thus are a baseline for precipitation in these mountains. PI created high resolution 3D hydrology model to simulate summertime flash floods in 3 headwater catchments, including the Cataloochee Creek basin (the other two outside the park). During landslide events, the pressure of water in soil pores on steep slopes overcomes the shear strength of the soils, causing the slopes to become unstable. Accurate spatial and temporal distributions of precipitation are key for the hydrological model, and prediction of extreme events in particular. The model being developed for rainfall within the mountains will allow for correcting errors of space and time in weather radar interpretation of precipitation events, making the prediction of landslide events more accurate."