Underground cave can be detected by Electrical Resistivity Tomography or ERT method (Zhou et al, 2000). Forward modeling was conducted to make spatial model of caves. Caves that were made by forward modeling were as cavities not as conduits. The spatial model of caves is shown on Figure 1.

Figure 1. Spatial model of caves
The model comprises of three main things such as cavities, water, and limestone. The true resistivity value of water would be 10-100 Ohm meter, and then for limestone would be 40-500 Ohm meter (Loke, 2000). The cavities would have different true resistivity values, the cavity that is fully filled with air would have very high resistivity value (the resistivity value of air was defined in 2000 Ohm meter). The cavity that filled with little amounts of water would have smaller resistivity value then the resistivity value would be getting smaller as the increasing of water amounts on the cavity. The resistivity model that represent the spatial model of caves is shown on Figure 2.

Figure 2. Resistivity model of cave's cavities
Simulation of DC current injection was conducted on the resistivity model by using Res2DMod software. Double Dipole configuration was chosen because of its advantages, better data resolution and less noises (Bernard et al, 2004).The result of apparent resistivity calculation from resistivity model is shown on Figure 3. Noises of apparent resistivity value from the calculation is assumed as 0 %. The real condition of caves environment actualy contributes high noise on the apparent resistivity calculation so 10 % noise value was put on the data to represent the real condition.

Figure 3. Apparent resistivity calculation from resistivity model
Res2DInv software was used for calculating true resistivity value from calculated apparent resistivity value of the model. Combination of Marquardt-Smoothness Constrained Least Square Inversion was chosen to get better result of resistivity pseudosection (Figure 4). Based on resistivity pseudosection, cavity that is fully filled with air would be easier to be detected by its high resistivity value, then cavity that is fully or mostly filled with water would be easier to be detected by its low resistivity value. The cavity that is filled with little amounts of water would be harder to be detected because the resistivity pattern on the resistivity pseudosection appears in disordered shape that not represent a cavity shape.

Figure 4. Resistivity pseudosection of cave's cavity
REFERENCES:
Bernard, Jean, & Orlando Leite, Fabrice Vermeersch. 2004. Multi-Electrode Resistivity Imaging for Environmental and Mining Application. Orleans: IRIS.
Loke, M. H. 2000. Electrical Imaging Survey for Environmental and Engineering Studies. 06/03/2009. http://www.geometrics.com.
Zhou, W., & B. F. Beck, J. B. Stephenson. 2000. Reability of Double-Dipole Electrical Resistivity Tomography for Defining Depth to Bedrock in Covered Karst Terranes. Environmental Geology, 39, 7.