Development of a prediction model of fuel moisture changes in a deciduous forest of Yeongdong Region in Korea
2014
Lee, S.Y. (Kangwon National University, Samcheok-Si, Gangwon-Do (Republic of Korea). Professional Graduate School of Disaster Prevention) | Kwon, C.G. | Ko, D.W. | Park, G.S. | Ohga, S.
Understanding how fuel moisture changes after a rainfall is important to predict forest fire risk, and knowing such change in advance can greatly assist in fire risk monitoring. To better understand the fuel moisture dynamics after a rainfall, we investigated how fuel moisture level changes across four different fuel layers (fall leaves, humus, top soil layer (<5 cm depth), and lower soil layer (5-10 cm depth)) under three different stand density levels (high, medium and low) after a significant rainfall event (>5 mm) during spring season. We measured variables including effective humidity, solar irradiation, wind speed, and days after rainfall. These variables were incorporated into developing a fuel moisture prediction regression model. Variables were measured daily for 6 days after a rainfall, for a total of 4 rainfall events in the spring of 2008 for model development, and one event in the spring of 2009 for model validation. Results show that in a low density stand, fuel moisture of the fallen leaves layer reached a dangerously dry level of 17% only after 3 days since rainfall, while at the medium and high density stands, fuel moisture level remained at 19-20% after 6 days since rainfall. Fuel moisture at the humus level was highest among all fuel layers, and remained at greater than 57% even after 6 days since rainfall. Top and lower soil layers both showed small to no changes in fuel moisture content throughout the sampling period. The prediction regression model showed a reasonable performance (R**2=0.56-0.90, p-value <0.001) and validated well against an independent set of measurements.
Mostrar más [+] Menos [-]Palabras clave de AGROVOC
Información bibliográfica
Este registro bibliográfico ha sido proporcionado por Agriculture, Forestry and Fisheries Research Information Technology Center