Reservoirs have many multi-functionalities that have an economic and sociological impact on society as a whole. These reservoirs have been used for daily water supply, energy production, food production, etc. Whether it is a natural or man-made reservoir, it has a direct impact on lives and has an increasing amount of importance for the growing population and expanding economy. Understanding the variation in the height of the water level in (large) reservoirs is important geotechnical data for many applications. This includes the volume of drinking water existing to supply a city; the amount of water available to irrigate crops downstream from a dam (Figure 1); the load exerted on a dam over time but also after an earthquake.

Figure 1: Shows an image of Rajasthan, India, taken through the SAR satellite image. The sentinel-1 radar signal is sensitive to ground moisture and can take images despite cloud covers, as SAR images can penetrate cloud covers. Therefore, it is possible to monitor water in reservoirs during any period of the year. On the left, water in reservoirs appears in dark blue while on the right side, during dry periods, the water level in reservoirs drops and it remains just the evidence of soil moisture (light blue color).

Given the number of reservoirs and the different sizes that these occupy, it is often difficult for administrations to have a precise and objective idea at all times of the water reserves stored in a given territory. For example, in Thailand, there are more than 4,000 reservoirs to monitor. It would be a challenging task to monitor all the reservoirs and extract the crucial water level information.

Skymap Global is working on a solution based on the processing of satellite images from the Sentinel constellation (Figure 2). Optical and radar images are combined to provide complementary sources of data. Cloud cover is an important limiting factor for the use of optical imagery. Therefore, in combination with radar, Skymap Global is developing a service available anytime, independent of cloud cover/ weather limitations. This is crucial for the extraction of information in a time series analysis. Together with Skymap Global Digital Elevation Models, the data is processed to provide an independent and reliable source of information to the decision-makers.

Figure 2: This is a high-level description of the method used to extract shoreline of reservoirs from optical and radar images. Once extracted, shorelines are carefully analyzed in a 3D environment to extract the most accurate elevation at a given moment. A report is delivered to the client showing the evolution of the water level through time.

Results
Figure 3 compares three images of the same reservoir. The water level dropped drastically from January to December 2019 interval. On the right side, a color composition indicated in turquoise color the ground surface discovered after the recession of the shoreline. Figure 4 shows a multitemporal color composition (below) made of three images of an abandoned meander used as a reservoir.

Skymap Global has the technical advances and methods to gain insights for SAR and optical data on water level monitoring in a large area. All of this can be done in our EOfactory platform. Results can be generated on a timely basis for decision-making and analysis. The power of analytics and machine learning on earth observations data is very important for reservoir management, making it possible for decision-makers to understand and make smarter decisions.