Name
Title | ||
|---|---|---|
Lunar Microwave Brightness Temperature: Model Interpretation and Inversion of Spaceborne Multifrequency Observations by a Neural Network Approach |
Abstract | ||
|---|---|---|
Understanding the lunar physical properties has been attracting the interest of scientists for many years. This paper is devoted to a numerical study on the capability of retrieving the thickness of the first layer of regolith as well as the temperature profile behavior from satellite-based multifrequency radiometers at frequencies ranging from 1 to 24 GHz. To this purpose, a forward thermal-electromagnetic numerical model, able to simulate the response of the lunar material in terms of upward brightness temperature (TB), has been used. The input parameters of the forward model have been set after a detailed investigation of the scientific literature and available measurements. Different choices of input parameters are possible, and their selection is carefully discussed. By exploiting a Monte Carlo approach to generate a synthetic data set of forward-model simulations, a physically based inversion methodology has been developed using a neural network technique. The latter has been designed to perform, from multifrequency TB's, the temperature estimation at the lunar surface, the discrimination of the subsurface material type, and the estimate of the near-surface regolith thickness. Results indicate that, within the simplified scenarios obtained by interposing strata of rock, ice, and regolith, the probability of detection of the presence of discontinuities beneath the lunar crust is on the order of 84%. The estimation uncertainty of the near-surface regolith thickness estimation ranges from 11 to 81 cm, whereas for the surface temperature, its estimation uncertainty ranges from about 1.5 K to 3 K, conditioned to the choice of radiometric frequencies and noise levels. |
Year | DOI | Venue |
|---|---|---|
2011 | 10.1109/TGRS.2011.2160351 | IEEE T. Geoscience and Remote Sensing |
Keywords | Field | DocType |
radiometry,microwave radiometry,first regolith layer thickness retrieval,lunar physical properties,inverse problem,monte carlo approach,astronomy computing,forward model simulations,lunar microwave brightness temperature,spaceborne multifrequency observation inversion,neural network (nn),forward thermal-electromagnetic numerical model,subsurface material type discrimination,inverse problems,moon microwave emission,satellite based multifrequency radiometers,neural network approach,radiative transfer (rt) model,temperature profile behavior retrieval,model interpretation,moon exploration,lunar surface,physically based inversion methodology,microwave measurement,monte carlo methods,astronomical spectra,lunar material response simulation,upward brightness temperature,neural nets,frequency 1 ghz to 24 ghz,monte carlo,surface temperature,estimation,moon,neural network,radiative transfer,artificial neural networks,integrated circuit,artificial neural network,probability of detection,materials,synthetic data,brightness temperature,physical properties | Monte Carlo method,Satellite,Brightness temperature,Classification of discontinuities,Regolith,Remote sensing,Radiometry,Synthetic data,Mathematics,Radiometer | Journal |
Volume | Issue | ISSN |
49 | 9 | 0196-2892 |
Citations | PageRank | References |
3 | 0.58 | 5 |
Authors | ||
5 | ||
Name | Order | Citations | PageRank |
|---|---|---|---|
| Mario Montopoli | 1 | 59 | 17.34 |
| Alessandro Di Carlofelice | 2 | 6 | 1.39 |
| Marco Cicchinelli | 3 | 3 | 0.58 |
| Piero Tognolatti | 4 | 10 | 2.09 |
| Frank S. Marzano | 5 | 41 | 15.92 |