roof is 16 by 27 inches. All parts are cut and 

 painted with outside white house paint prior 

 to nailing together. 



Hinges, hasp, and hasp loop are of brass. 

 The frame is nailed together with galvanized, 

 adhesive-coated sixpenny box nails, and the 

 smooth, two-surface, masonite sheeting is 

 nailed on with %-inch galvanized, flathead 

 roofing nails. 



The shelter is again painted after assem- 

 bling. Shelters are painted annually to keep 

 a bright, reflective, white surface. 



Construction Procedure 



The most economical method of construc- 

 tion is to make the shelters in groups of three 

 or multiples thereof. The pieces for the frame 

 of the shelter are precut and painted before 

 nailing together. 



Sawing is best done on a circular table 

 saw. The Presswood pieces should be drawn 

 on the large boards in such a way that the 

 first cut is a straight line across the middle 

 of the 4x8-foot sheet so that two pieces close 

 to the 4x4-foot size are produced. This aids 

 handling. Brace and leg pieces of one length 

 should all be cut at one time. For example, 

 for six shelters twelve 31 -inch legs should be 

 cut at the same setting. 



The end frames are nailed together first 

 and then the longitudinal crosspieces. The 

 frame pieces are precut from 1 by 2's. Legs 

 should be knotfree; but other pieces, such as 

 hasp blocks and hinge blocks, can have knots 

 in them since these are not strength pieces. 



The Use of the Instrument Shelter 



To prevent the shelter from being blown 

 over in high winds, four stakes are driven into 

 the ground next to the legs, the shelter is 

 leveled with a spirit level, and the legs are 

 nailed to the stakes. 



The location in which the shelter is placed 

 greatly influences the results obtained. For 

 comparing the effects of topography on local 

 climate, the vegetative site must be standard- 

 ized. A flat place with mowed grass away from 

 any trees or buildings is a good site. If the 

 effects of local vegetation on climate are to be 

 studied, the same topographic site should he 



used, with shelters placed in the various vege- 

 tations desired. As an example, the instru- 

 ments in a flat plane could compare the 

 weather under tree crowns in a forest open- 

 ing and an open meadow away from trees. 

 To determine the range of climate of a slope, 

 nine instruments would be used: three along 

 the top in various or replicated vegetative 

 covers, three similarly placed in the middle of 

 the slope, and three along the bottom of the 

 slope. The minimum period needed to make 

 reasonable climatic comparisons between 

 weather stations is 30 days. Measurements 

 should be repeated during the same periods 

 for at least 3 years, because wind and weather 

 patterns vary greatly from year to year. 



Accuracy of Weather Measurements 



Measurements with an especially accurate 

 platinum resistance thermometer showed 

 that the temperature inside the boxes varied 

 only about 0.2° C. from the temperature out- 

 side during the times of most rapid diurnal 

 temperature change. 



The standardized height of 18 inches 

 above the ground was selected in 1951 be- 

 cause this was found to be a common height 

 of blister rust cankers on white pine and of 

 blister rust infection on ribes. Conditions fa- 

 voring blister rust infections are extremely 

 rare at the height of a standard U.S. Weather 

 Bureau shelter. In rice disease studies, shel- 

 ter floors were set 3 inches above the water 

 surface because that was more nearly the 

 level of the disease infections on the plants. 



At 18 inches above the ground diurnal 

 variations in temperature were greater in 

 clear weather than for the standard U.S. 

 Weather Bureau shelter. Nightly minimum 

 temperatures were lower, and daytime maxi- 

 mum temperatures were higher. Humidities 

 were generally higher at the 18-inch level. 

 When placed at the height of a U.S. Weather 

 Bureau shelter with the bottom about 4V2 

 feet from the ground, the temperature and 

 relative humidity results were the same on 

 the same site. In cloudy weather, there were 

 almost no variations as a result of differences 

 in the height above the ground. 



July 1963/ 



2> < 

 E. P. VAN ARSDEL, 



Plant Pathologist 



