1: \begin{abstract}

2: We examine scatter and bias in weak lensing selected clusters,

3: employing both an analytic model of dark matter haloes and numerical

4: mock data of weak lensing cluster surveys. We pay special attention 

5: to effects of the diversity of dark matter distributions within

6: clusters. We find that peak heights of the lensing convergence map

7: correlates rather poorly with the virial mass of haloes. The

8: correlation is tighter for the spherical overdensity mass with a higher

9: mean interior density (e.g., $M_{1000}$). We examine the dependence of

10: the halo shape on the peak heights, and find that the root-mean-square

11: scatter caused by the halo diversity scales linearly with the peak

12: heights with the proportionality factor of $0.1-0.2$. The noise

13: originated from the halo shape is found to be comparable to the source

14: galaxy shape noise and the cosmic shear noise. We find the significant

15: halo orientation bias, i.e., weak lensing selected clusters on average

16: have their major axes aligned with the line-of-sight direction, and that

17: the orientation bias is stronger for higher signal-to-noise ratio

18: ($S/N$) peaks. We compute the orientation bias using an analytic

19: triaxial halo model and obtain results quite consistent with the

20: ray-tracing results. We develop a prescription to analytically compute

21: the number count of weak lensing peaks taking into account all the

22: main sources of scatters in peak heights. We find that the improved

23: analytic predictions agree well with the simulation results for high

24: $S/N$ peaks of $\nu\ga5$. We also compare the expected number count

25: with our weak lensing analysis results for 4~deg$^2$ of

26: Subaru/Suprime-Cam observations and find a good agreement. 

27: \end{abstract}

28: