is possibly more variable. Of interest is the close correlation of the array and control site at 30 cm depth (r = .988), indicating that the array may have little effect on diurnal soil temperature fluctuations at that depth, even though the control temperature is consistently higher. At the 5 cm depth, the correlation is not as evident (r = .923), since the control site appears to heat up rapidly as well as lose heat rapidly on cold days (e.g., Dec. 22). Further evidence of the ameliorating effect of the array is the higher correlation of 5 and 30 cm depth in the array (r = .956) than in the open desert (r = .855). Heat flux rates into and out of the soil are apparently significantly higher in the open desert than in the array as a whole. Figures A-3 and A-4 show soil temperatures within the collector array based on different microsites, as previously discussed, compared with the open desert. Of interest is the fact that at both depths the sunny gaps within the array have warmer soils than in the open desert. This is especially true at 30 cm, and supports our previous hypothesis concerning the possibility of microsites within the array receiving more intense radiation than would occur in the open desert. The shaded gaps are consistently and significantly the coolest soils at both depths. Why these soils are so much cooler than the soils immediately under the collectors is not obvious at the moment, but further studies may provide an explanation. A possibility is that more intense shading occurs there, as the soils underneath the collectors receive direct sunlight in the early morning and late afternoon hours. Soil Moisture The Colman blocks used for soil temperatures also give values that can be converted into the percent of soil moisture available. Therefore the same microsites were sampled for soil moisture as for temperature within the array (Fig. A-l) and at three control locations. 1977 was one of the driest years on record in Phoenix, and so the only significant rainfall event during the fall occurred on October 28-29, in which the average rainfall from the two nearest weather stations was 7.1 mm (0.28"). This rainfall event was not of sufficient magnitude to completely wet the dry desert soils except in front of the collectors where runoff down the panels
RkJQdWJsaXNoZXIy MTU5NjU0Mg==