My topic of research while in Michigan is on Land Use and Land Cover Change which made today's paper which I had to present especially interesting. The paper "Evaluating the effects of historical land cover change on summertime weather and climate in New Jersey: Land cover and surface energy budget changes" by Paul Wichansky et al (2008), investigates the sensitivity of a warm season climate of New Jersey to land cover change from 1880s versus 1990s. The areas of specific interest include the magnitude of land cover change and sensitivity on near surface air temperature, rainfall, dew point temperature, heat fluxes and climatological episodes.
Previous studies have shown that deforestation causes an increase in temperature and rainfall shifts, while reforestation causes cooling in the region. Urbanisation also causes an increase in temperatures and changes to the diurnal cycle. The interesting thing to note is that all three processes have occurred in the state of New Jersey within the last century. Reconstructing the land cover from historical topographical maps and land consensus, a full land cover map was available for direct comparison with NLCD data from 1992. Aggregated to 1km and combined into 8 vegetation classes, the researchers used RAMS and LEAF2 to compare atmospheric responses to land cover change.
They nested the grid in 3 levels (above-purple) and 3 ensemble members per land cover scenario, they used a 2 month period in June/July 1999. From NCEP data to force atmospheric boundary conditions, they used June as a spinup and July as a month for analysis. The soil moisture was described from observation.
The model comparisons showed an increase in near surface air temperature and dew point temperature in urban and deforested areas and cooler in reforested areas, just as previous research stated. The results overall come as no surprise but they way they use the ensemble without giving the spread or error is slightly weird.