Integrating Satellite Observation for Monitoring Inland Water Storage and Offering Database in Taiwan(1/3)

The proposed research aims to compute water level and associated discharge for inland water bodies inTaiwan, by using integrated remote sensing imagery and digital elevation model (DEM). The work includessurface type classification for river surface and adjacent land cover. Target areas include, from broad to finescale as time being, major reservoirs, major rivers, mid-small size lakes and narrow river/stream. Theremote sensing satellites utilized in this project include Landsat series, FORMOSAT-2, and SPOT series in8–30 meter temporal resolution and 1–16 day revisit cycle. The proof-of-concept study of this observationsystem, as demonstrated in mountainous area, has provided the water level estimate at 1-meter accuracy inevery 20 days, limited by Landsat revisit period and resolution. Further error analysis shows that thismethod with finer grain of input satellite imagery will yield better than 30 cm accuracy in every other day,for gradual terrain slope such as riverbanks and reservoirs in Taiwan. At the mean time, we will reprocessthe historical record of altimetry satellite data, e.g., TOPEX/Poseidon, Envisat, Jason-2, Cryosat-2,SARAL/AltiKa, and ICESat for river surface anomaly detection. During the Cal/Val stage we will useunmanned aerial vehicle (UAV) carrying high-resolution camera to obtain detailed classification map ofriver basin. Finally, we will offer an online database with graphical user interface (GUI) for users tocheck/download historical and near real-time water data at no-cost.