Introduction

The two most common methods of estimating Tau are performing a skydip and extrapolating from the CSO Tau monitor, which measures the zenith sky opacity at 225 GHz at a fixed azimuth. Tau can also be derived using a secant plot, but this requires a meaningful number of measurements of a bright source over a range of airmasses in very stable conditions to be accurate.

Skydips measure the sky brightness temperature over a range of elevations (usually between 80 and 15 degrees). When performing a skydip, SCUBA alternates between observing the sky, observing a hot load, and observing a cold load at each elevation. The physical temperature of the hot and cold loads can be measured, and are adjusted to match the bandwidth and central wavelengths of the various filters. Thus the loads serve to calibrate the observations of the sky itself. A model describing both the atmosphere (assuming a plane-parallel form) and the optical system is then fit to data in order to estimate the zenith sky opacity.

Ed Chapin (University of Victoria) analysed the skydip and CSO Tau data taken before May 1998, and derived the following correlations between Tau₈₅₀, Tau₄₅₀, and Tau_CSO (refer also to Figure 1):

Tau₈₅₀ = 4.3 * (Tau_CSO - 0.006)

Tau₄₅₀ = 25 * (Tau_CSO - 0.011)

**Figure 1:** CSO Tau relations for data taken prior to May 1998. The left plot shows the Tau₈₅₀-Tau_CSO correlation, the right plot shows the Tau₄₅₀-Tau_CSO correlation.
$\begin{figure} \centering\epsfig{file=echapin_850vscso_1.ps,height=8cm}\epsfig{file=echapin_450vscso_1.ps,height=8cm}\end{figure}$

These relations have been regarded as an essential tool for reducing data. In this document, we present a re-analysis of how best to determine the sky opacity, including a new interpretation of CSO Tau data and new measurements of the temperatures of the hot and cold loads used to calibrate skydips. Combining this information, we present revised CSO Tau relations.

Before proceeding, we need to mention the SCUBA Upgrade Project, which was completed on October 11, 1999. New blocking filters were installed which have better transmission than the old ones (~10% improvement at 850µm, ~20% improvement at 450µm). As before, they are designed to cut out extraneous radiation that would otherwise be detected by the bolometers. The sensitivity of all filters will have been affected by this change. In addition, two new wideband filters centred at 450µm and 850µm (450W:850W) were installed. Their transmission profiles are shown here . The narrowband predecessors (450N:850N) are still available, although at 450µm the new wideband filter is considerably more sensitive under all weather conditions.

We have re-analysed the data for the pre-upgrade 450N:850N narrowband filters, the post-upgrade 450N:850N narrowband filters, and for the new 450W:850W wideband filter system. Please note that so far we have only been able to re-analyse the 450-µm and 850-µm data. We have not yet tackled the 350-µm, 750-µm, 1350-µm or 2000-µm filters.