Dyson Sphere Search HistoryA number of searches for Dyson Spheres have been made in the past. A few candidates have been identified but discounted for various reasons.
Sagan and Walker carried out an early analysis of the possibility of detecting a Dyson Sphere. They showed that a search out to 1000 pc was feasible even with sixties technology but that the possible confusion with natural signatures could require searches for other artifacts of intelligence such as radio signals associated with a candidate source.
Jugaku and colleagues have carried out a series of searches for partial Dyson Spheres. Typically they used the 2.2 μm K band as an indicator of the photospheric radiation of a star hosting a partial Dyson Sphere and then looked for an infrared excess in the IRAS infrared satellite 12 μm band. A 1 magnitude difference would arise if the Dyson Sphere covered 1% of the host star. The measured differences were characteristically less than 0.3 magnitude which is consistent with measurement errors. They selected a set of 1774 stars from the Woolley catalog nearer than 25 pc and found 458 with matches in the 12 μm IRAS band. They looked at 384 of these stars for infrared excesses. With the exception of a few cases discussed in their 1990 article they found no sources with excesses suggestive of a partial Dyson Sphere covering as much as 1% of the host star.
Slysh and Timofeev at al. have the IRAS database for a different approach. Slysh investigated the flux at the maximum of a Dyson Sphere spectrum. He estimated that all Dyson Spheres with temperatures from 50 to 400 ēK within 1 kpc of the sun should have been detected. The Timoreev search looked at a population of IRAS sources in the 110-120 and 280-290 ēK temperature range as established by Kardashev and others and did Planck blackbody fits to the four IRAS bands. They fitted by minimizing to a Planck distribution. (Note that no Planck spectrum correction is made on the four measured fluxes from the filters.) Sylsh identified one possible Dyson Sphere candidate, G357.3-1.3. The Timofeev at al. search identified 10 or so candidates but ruled out most of them, often on the basis of associations.
More recently several other searches have been conducted for partial Dyson Spheres. Globus, Backman, and Witteborn have searched by looking for a temperature/luminosity anomaly due to the fact that the luminosity of a star surrounded by a partial Dyson sphere would be lowered compared to a naked star of the same temperature. Conroy and Werthimer have searched by constraining the Jugaku infrared excess technique to older stars using a list of 1000 nearby older stars compiled by Wright and Marcy. Using older stars eliminates thick dust clouds around young stars. They also correlate with the rich K band near-infrared ground based data from 2MASS. They have found 33 candidates in the 12 μm IRAS band with 3 σ excesses from the mean.
A 2009 investigation was carried out using IRAS data by me and and my colleagues.
In the 2015 time frame Jason Wright and his colleagues searched for large scale objects using the Wide-field Infrared Explorer data WISE.
Dyson Sphere Search Bibliography