ROUGH DRAFT authorea.com/62484

# Introduction

The goal of this document is to outline the deliverable products and performance requirements for the ZTF Data Systems. This is to ensure that ZTF can achieve the science goals described in the ZTF Science Requirements Document. The requirements listed below apply lessons learned from PTF and iPTF to specify what should be possible with ZTF. In this document we will try to review the current status, precision and accuracy, of the PTF pipeline, and suggest ways to improve it.

We also provide the list of required pipeline steps, meta data and data products to be stored. Some of this requirements may have implications to hardware apparatus (e.g., flat field screen?), and some to computing requirements and data storage.

The primary science goals for ZTF, as outlined in the science requirements document, are listed below:

1. Discovery of Fast, Young, and Rare Transients at near real time;

2. Searches for Electromagnetic Counterparts to Gravitational Wave Sources;

3. Producing high precision light curves for all sources;

4. Asteroid and streaks Discovery.

# Performance Requirements

$$\bigstar$$ Adam: The text here is confusing - not clear to Eran $$\bigstar$$

The performance requirements for ZTF are derived from the anticipated science, and the survey is designed first and foremost as a time-domain discovery engine. Broadly speaking, ZTF will focus on the study of two classes of sources: transients and variables. For the purposes of this document, transients are defined as sources that are visible for a limited duration of time as the brighten then fade (e.g., novae, supernovae [SNe], gamma-ray bursts [GRBs]), while variables are persistently detectable, but continuously changing brightness (e.g., RR Lyrae stars [RRL], eclipsing binaries [EBs], active galactic nuclei [AGN]).1

While there are some science cases that will not require extremely high precision or accuracy, most, if not virtually all, of the science conducted by ZTF will be limited by the photometric and astrometric accuracy of the survey data. Using previous wide-field surveys, including PTF, as a guide, we outline realistic goals for ZTF below.

1. There are numerous examples of sources that straddle the boundary between these two classes (e.g., a galactic nova), though these special cases do not affect the following discussion.

## Scope

The first requirements for the survey concern the scope of the observations. ZTF must survey the entire sky visible from Palomar Observatory. In practice, this will cover nearly the entire footprint of the Pan-STARRS1 (PS1) 3$$\pi$$ survey (Dec > -30$$^\circ$$). To enable the search for transients, it is essential that the full survey area be observed and reference images be created within the first year of the survey. Furthermore, a systematic grid of discrete fields should be developed prior to the start of the survey, and all observations should be confined to these fields. This will ensure that, once reference images are made, every frame taken by ZTF will be suitable for transient discovery. We recommend that the fields consist of two overlapping grids, offset by $$\sim$$half the length of the diagonal such that chip gaps can be filled and the corners of the FOV in one grid, which suffer the most from vignetting, correspond to the center of the FOV in the alternate grid. Alternative solutions may be appropriate, but they must be defined prior to the start of the survey.

Filter Set

An important lesson from PTF is that the survey itself must provide color information, if this is to be utilized in near-real time. While there are many $$\sim$$1-m class telescopes suitable for photometric follow-up of ZTF discoveries, unlike ZTF, these telescopes will in almost all cases lack reference images. Thus, with the exception of variable stars and “hostless” transients, these telescopes will need to wait for the transient to completely fade before obtaining the appropriate reference images for image subtraction. Thus, it is viewed as a requirement that ZTF, unlike PTF/iPTF, have at least 2 photometric filters in order to produce color information in quasi-real time.

Specification: the filters for ZTF are specified in Table \ref{tbl:ztf_filters}.

\label{tbl:ztf_filters} The photomet