The speed of gravitational waves from strong lensed gravitational wave-electromagnetic signals

We propose a new model-independent measurement strategy for the propagation speed of gravitational waves (GWs) based on strongly lensed GWs and their electromagnetic (EM) counterparts. This can be done in a two-fold way: by comparing arrival times of GWs and EM counterparts and by comparing the time delays between images seen in GWs and EM counterparts. The lensed GW-EM event is perhaps the best way to identify an EM counterpart. Conceptually this method does not rely on any specific theory of massive gravitons or modified gravity. Its differential setting (i.e. measuring the difference between time delays in GW and EM domains) - makes it robust against lens modeling details (photons and GWs travel in the same lensing potential) and against internal time delays between GW and EM emission acts. It requires, however, that the theory of gravity is metric and predicts gravitational lensing similar as General Relativity. We expect that such test will become possible in the era of third-generation gravitational-wave detectors, when about 10 lensed GW events would be observed each year. The power of this method is mainly limited by timing accuracy of the EM counterpart, which for kilonova is around $10^4$ sec. This uncertainty can be suppressed by a factor of $\sim 10^{10}$, if strongly lensed transients of much shorter-duration associated with the GW event can be identified. Candidates for such short transients include short gamma-ray burst and fast radio bursts.