Cosmological forecasts with the clustering of weak lensing peaks

Maximising the information that can be extracted from weak lensing measurements is a key goal for upcoming surveys such as LSST and Euclid. This is typically achieved through statistics that are complementary to the cosmic shear two-point correlation function, the most well established of which is the weak lensing peak abundance. In this work, we study the clustering of weak lensing peaks, and present parameter constraint forecasts for an LSST-like survey. We use the cosmoslics $w$CDM simulations to measure the peak two-point correlation function for a range of cosmological parameters, and use the simulation data to train a Gaussian process regression emulator which is applied to generate likelihood contours and provide parameter constraint forecasts from mock observations. We investigate the dependence of the peak two-point correlation function on the peak height, and find that the clustering of low amplitude peaks is complementary to that of high amplitude peaks. Consequently, their combination gives significantly tighter constraints than the clustering of high peaks alone. The peak two-point correlation function is significantly more sensitive to the cosmological parameters $h$ and $w_0$ than the peak abundance, and when the probes are combined, constraints on $Ω_{\rm m}$, $S_8$, $h$ and $w_0$ improve by at least a factor of two, relative to the peak abundance alone. Finally, we compare the forecasts for weak lensing peaks and weak lensing voids, and show that the two are also complementary; both probes can offer better constraints on $S_8$ and $w_0$ than the shear correlation function by roughly a factor of two.