Old isolated neutron stars have sparked the imagination of astronomers since the early days of X-ray astronomy (Ostriker et al. 1970; Helfand et al. 1980). This quest was reignited by the launch of ROSAT and led to several theory papers estimating the detectability of old neutron stars in the ROSAT All-Sky Survey. A general description of the ROSAT satellite is presented in (Trümper 1983; Pfeffermann et al. 1986) and the All-Sky Survey was described by Voges (1992).
Treves & Colpi (1991) were the first to calculate in detail the probability of detecting old neutron stars with ROSAT. Later Blaes & Madau (1993) published a more extensive study. Both groups arrived at the similar conclusion of a high probability to detect old neutron stars. When preliminary results from the Sky-Survey showed that these estimated were too optimistic, both groups reevaluated their calculations (Madau & Blaes 1994; Colpi et al. 1993) arriving at lower numbers.
A comparison of the pulsar birth rate and the number of known pulsars shows that we have strong indications to believe that the vast majority of neutron stars remains undetected, although more than 650 radio pulsars have been discovered. Only a small number of neutron stars, not active as radio pulsars, are detected as high energy sources. Almost all of which are in close binary systems. Where are the missing old isolated neutron stars?
At birth, a neutron star starts to cool quickly through neutrino emission. Through this process and continuous cooling through photon emission, most of the initial heat is dissipated after less than a million years. Once its surface temperature drops below 105K, a neutron star is undetectable to most instruments. However, the surface of a neutron star might be heated through accretion from the interstellar medium even after the initial thermal energy content is depleted. The resulting surface temperature depends strongly on the accretion geometry but it is generally expected to be on the order of 106K. Most estimates have assumed that the emergent surface spectrum is a black-body spectrum. Some theorists have speculated that the surface emission might be harder than a black-body spectrum, see e.g. Zampieri et al. (1995), and have calculated the effect of this assumption on the X-ray background (Zane et al. 1995). However, all calculations confirm that the emission spectrum will not deviate more than 40 percent from a blackbody spectrum. Therefore, because of the limited energy resolution available with ROSAT, the spectrum will not be noticeably different from a Planck spectrum. For this reason I will use a simple single-temperature Planck spectrum for all my estimates.
Based on pulsar birth statistics, the total number of neutron stars in our
Galaxy is estimated to be 
. The local neutron star
density was derived to be 
with a half density scale height of 280 pc (Madau &
Blaes 1994). A neutron star may be found as close as ten parsecs.
Today, only three candidates for old neutron stars accreting from
the interstellar medium have been proposed (Stocke et al. 1995;
Walter et al. 1996; Haberl et al. 1996). All three
candidates will require further work before they are unambiguously
identified. A recently published search for old neutron stars
(Belloni et al. 1997) has turned up four less well established
candidates. Meanwhile, systematic identification of ROSAT sources in the
galactic plane (Motch et al. 1997a,b) has sufficiently
progressed to demonstrate that the number of candidate objects
for old neutron stars is much smaller than expected earlier.
This conflict between the number of old neutron stars predicted and detected motivated me to search the ROSAT All-Sky survey systematically focusing of the regions with the highest expected neutron star densities. The current paper presents the results from a comprehensive identification program of X-ray sources in molecular clouds at high galactic latitude. Paper II covers sources in galactic dark clouds. Here, I present the X-ray source sample and its correlation with the astronomical catalogs SIMBAD and NED. I include finding charts, optical imaging and spectroscopic observations and the results from radio observations of all sources at the VLA. This paper describes the properties of the source sample and the observational data. A later, third, paper will discuss in detail the implications of the surveys in Papers I and II.