Search

Prediction of ionospheric parameters, such as ionospheric F2 layer critical frequency (foF2) at low latitude regions is of significant interest in understanding ionospheric variation effects on high-frequency communication and global navigation satellite system. Currently, deep learning algorithms have made a striking accomplishment in capturing ionospheric variability. In this paper, we use the state-of-the-art hybrid neural network combined with a quantile mechanism to predict foF2 parameter variations under low and high solar activity years (solar cycle-24) and space weather events. The hybrid neural network is composed of a convolutional neural network (CNN) and bidirectional long short-term memory (BiLSTM), in which CNN and BiLSTM networks extracted spatial and temporal features of ionospheric variation, respectively. The proposed method was trained and tested on 5 years (2009–2014) of ionospheric foF2 observation data from Advanced Digital Ionosonde located in Brisbane, Australia (27°53′S, 152°92′E). It is evident from the results that the proposed model performs better than International Reference Ionosphere 2016 (IRI-2016), long short-term memory (LSTM), and BiLSTM ionospheric prediction models. The proposed model extensively captured the variation in ionospheric foF2 feature, and better predicted it under two significant space weather events (29 September 2011 and 22 July 2012). [ABSTRACT FROM AUTHOR]

In the Australian summer season of 2022, exceptional rainfall events occurred in Southeast Queensland and parts of New South Wales, leading to extensive flooding of rural and urban areas. Here, we map the extent of flooding in the city of Brisbane and evaluate the change in electricity usage as a proxy for flood impact using VIIRS nighttime brightness imagery. Scanning a wide range of possible sensors, we used pre-flood and peak-flood PlanetScope imagery to map the inundated areas, using a new spectral index we developed, the Normalized Difference Inundation Index (NDII), which is based on changes in the NIR reflectance due to sediment-laden flood waters. We compared the Capella-Space X-band/HH imaging radar data captured at peak-flood date to the PlanetScope-derived mapping of the inundated areas. We found that in the Capella-Space image, significant flooded areas identified in PlanetScope imagery were omitted. These omission errors may be partly explained by the use of a single-date radar image, by the X-band, which is partly scattered by tree canopy, and by the SAR look angle under which flooded streets may be blocked from the view of the satellite. Using VIIRS nightly imagery, we were able to identify grid cells where electricity usage was impacted due to the floods. These changes in nighttime brightness matched both the inundated areas mapped via PlanetScope data as well as areas corresponding with decreased electricity loads reported by the regional electricity supplier. Altogether we demonstrate that using a variety of optical and radar sensors, as well as nighttime and daytime sensors, enable us to overcome data gaps and better understand the impact of flood events. We also emphasize the importance of high temporal revisit times (at least twice daily) to more accurately monitor flood events. [ABSTRACT FROM AUTHOR]