RISERnet networks captured fine-grained spatial and temporal data about the environmental conditions in the forest. One example of how such sensed data can be useful in understanding changes in the environment, is in monitoring fuel moisture content in the field. Information about fuel moisture content (FMC) is important because it strongly affects fire behavior. Fuel moisture is also closely related to several other widely used measures of, including drought factor and KBDI (Keetch-Byram Drought Index). MoistuRISER will develop the capability to use data from sensors, such as those deployed in RISERnet, to provide better information about fuel moisture conditions.
FMC is primarily a function of temperature, relative humidity, wind speed, and radiation. All four of these parameters are measured by on-board sensors in RISERnet. Consequently, using established physical and empirical relationships, it should be possible to compute FMC measures based on RISERnet sensed data. In addition, the motes in the Powelltown network are armed with contact sensors of fuel moisture content. This offers a novel scenario for measuring fuel moisture content.
MoistuRISER has four key stages:
-
Fuel moisture measurement in the field.
Characterization: Before attempting to model fuel moisture directly, it is important to understand the spatial and temporal characteristics of the RISERnet data across our study area in Olinda. Thus, work is looking closely at the data, in particular better understanding the statistical and error characteristics of RISERnet temperature, relative humidity, wind speed, and radiation data.
- Modeling: Having thoroughly understood the data characteristics, stage 2 aims to adapt existing physical and empirical models of fuel moisture (such as those of McArthur and Matthews) to compute fuel moisture based on RISERnet data.
- Validation: In order to test whether the computed FMC is an accurate reflection of the actual fuel moisture conditions, stage 3 will validate the computations by comparison with field observations of fuel moisture. Further, we will compare the accuracy of model-based and contact-sensor-based scenarios. It will be also investigated whether more accurate estimates of FMC can be obtained by combining the weather data and direct measurements from the fuel layer.
- Extrapolation: Finally, in stage 4 we will attempt to understand the relationships and structures in the FMC computation, in particular to look for correlations with other variables, such as slope, aspect, and vegetation type, in an attempt to extrapolate from over larger spatial and/or temporal extents.
