This is the definition adopted by the PDS Small Bodies Node for the format that should be used for shape models of small bodies. The model of the nucleus of comet Hartley 2 follows this format. General description of the model Shape models are defined using a set of points distributed across the surface of the body. These points, or vertices, are linked together in threes to form triangular plates or facets that approximate the local surface topography. Each vertex is incorporated into as many plates as necessary to define the shape of the body (or at least the parts of it that have been modeled). The total number of vertices and plates that are used is dependent on the spatial resolution that is available for defining the shape of the body. Coordinate systems The vertex positions are defined in a body-centered system using either cartesian coordinates or planetocentric spherical coordinates. Spherical coordinate systems should follow the IAU definition for small bodies. By this definition, the positive pole lies in the direction of the rotational angular momentum vector, and positive latitudes increase in that direction. Longitudes should be measured from 0 to 360 degrees in a right-hand coordinate system from a designated prime meridian. Whenever possible, the cartesian coordinate system should maintain a logical relationship to the rotational reference frame (e.g., +Z corresponds to the positive spin axis, +X corresponds to the prime meridian). If the rotational properties are not known, then a description of the adopted coordinate system should be included to allow for a transformation to a different system if the rotational properties are derived in the future. Model format The format for the model is defined to be an ascii file that is divided into three different, consecutive tables. Table 1 consists of a single line with two columns: Column 1 -- Total number of vertices that are included for defining the surface (N_vert) Column 2 -- Total number of plates that are used to represent the body's surface (N_plate) Table 2 consists of N_vert lines, each of which defines the position of one vertex. Each line contains three or more columns: Column 1-3 -- Position of a vertex in the adopted coordinate system. Column 4-n -- (Optional) Additional information about the vertex. (e.g., a flag denoting a point derived from observational data vs. an unobserved point that has been filled in by assuming a triaxial ellipsoid.) Table 3 consists of N_plate lines, each of which identifies three vertices from Table 2. The three vertices are connected to form a triangular plate on the body's surface. Each line contains three or more columns: Column 1-3 -- Vertex numbers denoting three vertices that are connected to form a plate. Each number is the offset from the first line from Table 2 (0 is the first vertex, 1 is the second, etc.). The vertices should be listed in an order such that when connected in the direction defined by the right hand rule, the resulting surface normal points to the exterior of the body. Column 4-n -- (Optional) Additional information about the plate. (e.g., a flag denoting that the plate contains vertices derived from observational data vs. those from an assumed triaxial ellipsoid. The column may also contain "-1" which denotes the end of a triangle in the VRML format.) If the entire surface has been modeled, the connectivity should be defined such that when the plates are assembled, the body forms a closed surface. For example, longitude 359 deg should connect back to the previously defined point at 0 deg, rather than to a mirrored point at 360 deg, and all of the plates that have a corner at a pole should connect to a single vertex at latitude (+/-)90 deg.