For example, by simulating hyperspectral data with different spatial resolution, Luo  evaluated the adaptability of linear spectral unmixing to different levels of spatial resolution. Jiao  simulated hyperspectral data to evaluate the influence of spatial and spectral resolution to vegetation classification. In addition, simulated data are often used to evaluate and test novel algorithms such as target detection and identification algorithms in hyperspectral remote sensing. There is no easy method to simulate hyperspectral data for testing the performance of these algorithms. If simulated hyperspectral data can be easily obtained, it will greatly help the testing and development of new algorithms.
The universal pattern decomposition method (UPDM) is a sensor-independent method which can be considered as a spectral reconstruction approach, in which each satellite pixel is expressed as the linear sum of fixed, standard spectral patterns for water, vegetation, and soil, and the same normalized spectral patterns can be used for different solar-reflected spectral satellite sensors . Sensor independence requires that analysis results for the same sample are the same or nearly the same regardless of the sensor used. Based on this trait, here we present a method based on the universal pattern decomposition method (UPDM) to achieve the goal of simulating hyperspectral data from multispectral data, which can be considered either a method of spectral construction or spectral transform. The hyperspectral and multispectral data are NASA EO-1 satellite/Hyperion and EO-1/ALI data, respectively (see Section 3.
2 for a brief introduction). First, we obtained ALI and Hyperion data covering the same area and AV-951 performed atmospheric correction to obtain surface reflectance data; here Hyperion data served as standard or real data to evaluate the results in the subsequent analysis. Then, we obtained the decomposition coefficients thought to be sensor-independent for the same sample by applying the UPDM to ALI data; these coefficients were subsequently used to construct Hyperion data. Before performing UPDM, standard pattern matrices of both sensors were calculated based on the standard spectral patterns (see Section 2 for details). Finally, the simulated Hyperion data were compared with the real Hyperion data, i.e., test data, to evaluate and assess this method.2.?Spectral Reconstruction Approach2.1. Review of the Universal Pattern Decomposition Method (UPDM)The spectral reconstruction approach is based on the UPDM, which is a sensor-independent method derived from PDM that has been successfully applied in many studies [14�C21].