Procedure for correcting 30 year maximum +ACY- minimum temperature sets -------------------------------------------------------------------------------- The following steps were taken to determine the functions linking monthly temperature differences and canopy coverages in the HJ Andrews. From them, radiation corrections were made to the 30-year temperature sets. 1. Determine suitable site pairs. Since elevation is one of the most important determinants of temperature at the scale of the HJ Andrews, pairing of similar-elevation sites was first made. +-50 meters was determined to be the window for acceptable pairs, and no sites over streams (thermograph and gaging station sites) were included because of cold-air drainage and other microclimatic issues that stream sites are often susceptible to. Of the available climate stations in the HJ Andrews providing reliable temperature data, there were 39 such pairs. Next, certain pairs were eliminated if one or both of the sites were known to have unusual temperature trends or possibly unreliable data or canopy images. For example, it was known at this point that both RS38 and RS89 have unusually high historical temperature values, so any pair containing one of these was eliminated. PRIMET and CS2MET both exhibit unusual long-term temperature trends which is likely a microclimatic effect in the bottom of the lower Lookout Creek Valley. The five sites from Bob Griffith's study were not used due to the unreliability of the site locations and hence the canopy images. RS12, though not a stream site, was determined to be too close to Lookout Creek to be free from cold-air drainage issues and not used. Of the remaining pairs, the final determination on whether to use them was based on the nature of their sites' canopies. Only pairs with both sites having continuous or discontinuous canopy types were used. The ideal pair was one which had a large radiation difference between the sites (one open, one closed), because sites with similar radiation differences tended not to provide conclusive trends with maximum temperature differences. Half of the site pairs used are of this type and consequently have a MET site as one of the pair, since these are the most open sites in the HJ Andrews. The final set of pairs: PAIR CANOPY CHARACTERISTICS ---- ---------------------- 1. VANMET/UPLMET Continuous Open/Continuous Open 2. UPLMET/RS04 Continuous Open/Continuous Closed 3. H15MET/RS05 Continuous Open/Continuous Closed 4. RS01/RS02 Discontinuous Open/Discontinuous Open 5. RS17/RS07 Continuous Closed/Continuous Closed 6. RS86/RS10 Continuous Open/Continuous Closed 7. VANMET/RS04 Continuous Open/Continuous Closed 8. RS05/RS03 Continuous Closed/Continuous Closed 2. Determine correction factors for monthly temperatures. Instead of looking at the relationship between insolation differences and temperature, we looked at normalized radiation differences and temperature. The difference between total monthly insolation (direct and diffuse) in a site pair was divided by the total monthly insolation value on a flat plain at an the average elevation of the HJ Andrews (843m). This is the value of direct and diffuse solar radiation upon a flat open surface with no topography or canopy shading, using theoretical IPW values. Monthly cloud effects and normal atmospheric attentuation were included in calculating these values. For each pair, these normalized differences in total solar radiation were plotted against differences in monthly maximum temperatures. The average slopes for each month: TMAX: JAN: 4.478 APR: 5.536 JUL: 5.101 OCT: 5.071 FEB: 4.998 MAY: 4.767 AUG: 5.549 NOV: 3.554 MAR: 5.064 JUN: 5.377 SEP: 5.771 DEC: 4.666 TMIN: JAN: -0.091 APR: -0.082 JUL: -1.957 OCT: -0.605 FEB: 0.080 MAY: -0.668 AUG: -1.650 NOV: -0.622 MAR: -0.381 JUN: -0.539 SEP: -1.485 DEC: -0.010 Since there appear to be seasonal trends in the correction factors for both maximum and minimum temperatures, these monthly correction factors were used instead of a yearly average. We also computed and plotted normalized direct radiation differences and temperature differences, with no appreciable improvement in the R+AF4-2 value of the monthly trendlines. It is interesting to note the seasonal differences in the nature of the trendlines, particularly as they relate to cloudiness. The difference in maximum and minimum temperatures decreases during cloudy months. In the case of maximum temperatures, this is because during the sunniest months, direct radiation plays a much more significant role in the radiation regime of a site. During these sunny, clear months (July, August, and September), canopy plays a more significant role in determining minimum temperatures because of the loss of thermal radiation is dramatically reduced in a forest compared to an open clearing. It should be noted that certain physical site characteristics originally thought to have a significant effect on the function between maximum temperature differences and canopy differences were found to be insignificant in this study. For example, aspects and slopes of a pair did not seem to be as important a factor as the nature of their canopies. It appears then, that forest canopy plays a more significant role in a location's radiation regime than its slope or aspect. 3. Apply corrections and generate new temperature set. After the monthly correction factors were determined for maximum and minimum temperatures, the new temperature sets were made. Since the function determined relates normalized canopy differences to temperature differences, both the maximum and minimum temperature corrections were made to correct temperatures as if the sites themselves were on flat, open slopes with no radiation attenuation other than cloudiness and normal atmospheric attenuation. In order for PRISM to best estimate temperatures on pixels in between sites, this correction type was deemed most suitable. canopy+AF8-temp+AF8-corr+AF8-procedure.txt 12/01