RAFAEL: Rainfall estimation using signalling data of satellite communication network

Rain has a significant influence on the global economy. It influences agriculture and the success of harvest, creating inflationary/ deflationary effect on agricultural produce and societal consumption of the same. It creates river and flash floods, which may vanish homes and businesses causing loss of life and property. It can interrupt traffic connections, supplies for electricity, gas, water and medical care bringing economic activity to a standstill.  Apart from this indirect economic impact, rain strongly influences the insurance business. Accurate estimation of rainfall would allow insurance companies to calculate the premiums precisely. The measurement of rainfall might appear to be straightfor­ward at first; however, it is highly variable spatially and temporally, making it very difficult to measure satisfactorily. Therefore, denser measurement networks are required in order to capture this variability.

However, due to high operational cost of such infrastructure, only 1,675 rain gauges can be found across 10 million square kilometres of Europe and around 8000 worldwide with satisfactory temporal resolution. There are other solutions for rainfall measurement: rain radars and Infrared Satellite Imagery. Although they have advantage of higher coverage area, but they suffer from high cost and poor temporal accuracy. This motivates the need for a cheap, accurate and innovative technology for rainfall measurement.

RAFAEL (Rainfall estimation using signalling data of satellite communication network) develops cheap, accurate and innovative technology for rainfall measurement. The main idea of RAFAEL is to extract rainfall information from the loss in signal strength between satellites and satellite ground terminals by employing advanced signal processing and machine learning techniques. There are more than 300,000 satellite ground terminals across Europe and 2 million worldwide; these can be transformed into reliable and real time rainfall measurement sensors. The signalling data is already aggregated into a single database, eliminating the need to create an elaborate network to feedback such sensor data. Further, replenishment/ installation of satellite ground terminals does not add to the database generation cost. Thus, the proposed idea of using signalling data is an efficient alternative to other database generation methods in terms of accuracy and CAPEX/ OPEX.

This Proof of Concept project builds on the results of the cooperation between SES and the University of Luxembourg/SnT. The aim is to valorise these results into minimum viable products that deliver reliable and cost effective rain estimation with high resolution targeting among others, the insurance and agriculture sector. 

The main idea is to use signalling data of the satellite communication links to estimate the rainfall. In a typical satellite communication system, content to/ from remote satellite ground terminals are aggregated at the gateway stations. Such stations are operated and maintained typically by satellite operators (like SES). The communication is bidirectional between the gateway stations and satellite ground terminals in broadband interactive satellite services (such as internet provisioning).

To maintain a certain level of quality of service for users, gateway stations continuously monitor the links between satellite and satellite ground terminals (enabled by the bidirectional link). Signal to noise ratio (S/N) parameter denotes quality of the received signal and the link. Satellite ground terminals measure this parameter and send it back to the gateway station. The S/N parameter is highly dependent on the link condition and is mainly affected by the rain attenuation at the operational Ka/band frequencies (broadband satellite). Therefore, there is clear correlation between the S/N measurements and amount of the rainfall. By using appropriate signal processing and machine learning techniques, we are able to estimate the rainfall from S/N measurements to a high accuracy.   

RAFAEL’s products have the following novel and disruptive aspects, which are not shared with the existing solutions:

• Low Cost:One of the main advantages of the RAFAEL compared to the existing solutions in the market is its low cost. The data collection part does not require any new investment in terms of equipment, interfaces and protocols. S/N data is already available at a centralized gateway station and ready to use after minimal processing. The CAPEX and OPEX are low.
• Large Coverage and Large Number of Terminals:With a single Geostationary (GEO) satellite, it is possible to cover the whole of Europe, North Africa and Middle East. In addition, there exist large number of terminals in the broadband satellite networks and it is possible to have access to their data at gateway stations. Figure ‎1‑4 shows the number of broadband satellite service consumers (terminals) worldwide. As can be seen, there are more than 2.5 million consumers currently with about 7 million consumers predicted by 2025. It can also be seen that number of consumers across Europe is around 300,000. These users are mainly served by three satellite operators in Europe: SES, Eutelsat and Avanti. In the US, ViaSat is one of the main broadband service provider with more than one million users.
• Real Time Data:In this considered satellite system, the S/N data is available at gateway every few minutes; such a temporal resolution will be more than sufficient for considered applications.
• Accuracy:Theconsidered satellite MW links operate at Ka-band, which is suitable to observe rainfall attenuation due to its sensitivity. It is known that rainfall estimation using MW links have better accuracy than rain radar and comparable to accuracy of the rain gauge. Our initial results confirm this accuracy.

Contact

Bhavani Shankar