We estimate the energy spectra of precipitating electrons from optical emissions and cosmic noise absorption (CNA) that were observed with the all-sky imager (ASI) and the 16 ¿ 16-element imaging riometer at Poker Flat Research Range (PFRR) (65.11¿ N, 147.42¿ W), Alaska. Maxwellian energy parameters (peak energy and total energy flux) are derived from the auroral spectroscopic emissions according to the conventional photometric method. We theoretically estimate CNA from the derived Maxwellian spectra and compare it with the observed CNA. The difference between the estimated CNA and the observed (δCNA) is used as an indicator of the difference in high-energy electron flux (E > ~20 keV) between the actual energy distribution and Maxwellian. This analysis is carried out for two substorm events in the evening sector to show the temporal and spatial variations of the energy spectra. Event 1 exhibits two arcs of CNA during the growth phase of the substorm. The results for this event suggest that the energy spectra of the two CNA arcs have different shapes, and this difference is revealed in the precipitating electron flux quasi-simultaneously measured by NOAA-17. Event 2 is a typical substorm that consists of growth, expansion, and recovery phases. The δCNA changes throughout the course of the substorm, which can be consistently explained by the energy spectrum variation of the precipitating electrons measured in previous studies. Furthermore, the energy spectra consistent for both optical and CNA data are estimated by assuming a kappa or a Gaussian distribution. This estimation method based on ground observations of the energy spectra has the potential to provide significant information on substorms.