Illumination and Temperature on Rough Terrain: Fast Methods for Solving the Radiosity Equation
Abstract
Illumination and the thermal environment are key to understanding the evolution of planetary surfaces and the distribution and transport of volatiles. With the increasing availability of large and highly accurate shape models collected by recent missions (e.g., MRO, LRO), thermal simulations in which all shape elements (triangles) may interact with each other are challenging computational problems. We aim to solve such problems posed on global or large and highly refined domains. We conduct these large-scale thermal simulations using the radiosity method, a finite element method which has seen extensive use in computer graphics. The complexity of the radiosity method normally scales quadratically with the number of elements for irregular geometry, which would make it intractable for large, detailed regions. We therefore apply hierarchical algorithms based on low-rank matrix approximation and geometric clustering to scale these algorithms so that huge problems can be solved on individual workstations. To investigate numerical convergence, we compute the irradiance on a hemispherical cap, a problem which has an analytic solution. Additionally, to stress test our solver, we apply it to a large shape model of 67P/Churyumov-Gerasimenko obtained from Rosetta and to LRO NAC digital elevation maps (DEMs) of polar lunar craters hosting volatiles.
- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2019
- Bibcode:
- 2019AGUFM.P31C3476E
- Keywords:
-
- 6205 Asteroids;
- PLANETARY SCIENCES: SOLAR SYSTEM OBJECTS;
- 6230 Martian satellites;
- PLANETARY SCIENCES: SOLAR SYSTEM OBJECTS;
- 6250 Moon;
- PLANETARY SCIENCES: SOLAR SYSTEM OBJECTS