# Dark Energy Survey Year 1 results: Validation of weak lensing cluster member contamination estimates from P(z) decomposition

## TL;DR

This paper validates a $P(z)$ decomposition method to accurately estimate and mitigate cluster member contamination in weak lensing galaxy catalogs, improving galaxy cluster mass calibration accuracy in the Dark Energy Survey Year 1 data.

## Contribution

The study demonstrates the effectiveness of the $P(z)$ decomposition technique through simulations and real data, showing it outperforms correlation-based methods in estimating contamination.

## Key findings

- $P(z)$ decomposition accurately estimates cluster member contamination.
- Simulation results show excellent agreement with true contamination profiles.
- Method improves cluster mass calibration precision in weak lensing studies.

## Abstract

Weak lensing source galaxy catalogs used in estimating the masses of galaxy clusters can be heavily contaminated by cluster members, prohibiting accurate mass calibration. In this study we test the performance of an estimator for the extent of cluster member contamination based on decomposing the photometric redshift $P(z)$ of source galaxies into contaminating and background components. We perform a full scale mock analysis on a simulated sky survey approximately mirroring the observational properties of the Dark Energy Survey Year One observations (DES Y1), and find excellent agreement between the true number profile of contaminating cluster member galaxies in the simulation and the estimated one. We further apply the method to estimate the cluster member contamination for the DES Y1 redMaPPer cluster mass calibration analysis, and compare the results to an alternative approach based on the angular correlation of weak lensing source galaxies. We find indications that the correlation based estimates are biased by the selection of the weak lensing sources in the cluster vicinity, which does not strongly impact the $P(z)$ decomposition method. Collectively, these benchmarks demonstrate the strength of the $P(z)$ decomposition method in alleviating membership contamination and enabling highly accurate cluster weak lensing studies without broad exclusion of source galaxies, thereby improving the total constraining power of cluster mass calibration via weak lensing.