Channel changes from 1919 to 1989 were documented in two study reaches of the Merced River in Yosemite National Park through a review of historical photographs and documents and a comparison of survey data. Bank erosion was prevalent and channel width increased an average of 27% in the upstream reach, where human use was concentrated. Here, trampling of the banks and riparian vegetation was common, and banks eroded on straight stretches as frequently as on meander bends. Six bridges in the upper reach constrict the channel by an average of 38% of the original width, causing severe erosion. In the downstream control reach, where human use was minimal, channel widths both decreased and increased, with a mean increase of only 4% since 1919. Bank erosion in the control reach occurred primarily on meander bends, The control reach also had denser stands of riparian vegetation and a higher frequency of large woody debris in channels. There is only one bridge in the lower reach, located at the downstream end. Since 1919, bank erosion in the impacted upstream reach contributed a significant amount of sediment (74,800 tonnes, equivalent to 2.0 t/km2/yr) to the river. An analysis of 75 years of precipitation and hydrologic records showed no trends responsible for bank erosion in the upper reach. Sediment input to the upper reach has not changed significantly during the study period. Floodplain soils are sandy, with low cohesion and are easily detached by lateral erosion. The degree of channel widening was positively correlated with the percentage of bare ground on the streambanks and low bank stability ratings. Low bank stability ratings were, in turn, strongly associated with high human use areas. Channel widening and bank erosion in the upper reach were due primarily to destruction of riparian vegetation by human trampling and the effect of bridge constrictions on high flow, and secondarily to poorly installed channel revetments. Several specific recommendations for river restoration were provided to park management.

This paper describes progress in the development of a 25 ppm NOx combustor that requires no diluent injection or post-combustion treatment The combustor will be retrofittable in all existing Allison Model 501-K series industrial engines. The approach undertaken is based on lean-premix combustion design incorporating an efficient fuel and air pre-mixing, fuel staging, and advanced wall cooling. Extensive use has been made of Computational Combustor Dynamics (CCD) codes in the design of the low NOx combustor. Experimental work in support of the present effort includes atmospheric bench scale testing and high pressure rig testing. The bench tests have been performed to evaluate several candidate designs, to gain better understanding of general lean pre-mixed combustor behavior, and to verify model predictions. The bench test results have indicated good fuel/air mixing performance of the lean premixing domes. The high pressure simulated engine rig tests of the dry lean pre-mixed low emissions combustors using natural gas have demonstrated NOx levels less than 15 ppm vd (15% O2 corrected), well below the program goals.