Hubbard Brook Climate Reconstruction 1894-2000
Historical runs with the PnET model require a climate record that extends back through the period of know historical disturbances in the Hubbard Brook valley. Previously we have used mean data for years before 1956, or repeated the 1956-1993 record several times. As it would be preferable to use something closer to the real climate series, we have used the VEMAP II data set to extend the measured record back in time. VEMAP II used historical records to compile an estimated climate series for the contiguous 48 states on a 0.5x0.5 degree grid (http://www.cgd.ucar.edu/vemap/V2.html).
In the following dataset, all data from the period of historical measurement at HB are taken directly from the Hubbard Brook data set as posted on the website. In our recent update of 1998, 1999 and 2000 data, there were some missing value fill-ins provided by HB staff that are soon to be complete at the website. The monthly mean was computed for 1998, 1999 and 2000 data for Watershed 6 for temperature and precipitation, and the radiation data from the sensors at Headquarters. We also added monthly Nitrogen deposition data divided as NH4 and NO3 sources (in g/m2) from CASTNET (http://www.epa.gov/castnet/data.html) and NADP (http://nadp.sws.uiuc.edu) data. Ozone data from CASTNET was also added as an 11 year average value. The nitrogen is represented as a measured monthly varying value from 1978 to 1997, and as an average outside of those years. For use in PnET-CN, Nitrogen from 1900 to 1978 show a trend of .1 to 1 fraction of current values, as suggested by the EPA report EPA-454/R-00-002 “National Air Pollutant Emission Trends 1900-2000”. The year 1894 was added at the start of the Hubbard Brook climate file to provide an average year record for use in historic extrapolation before 1894 for the PnET model. The historic Hubbard Brook measured temperature and precipitation record begins in 1956. For prior years, the data are a function of the VEMAP II data set in the following way:
All VEMAP data were taken from the pixel located at 44.25N, 71.75W
All data types were extrapolated by first developing linear regressions between the measured data from and the VEMAP II data for the same period. The regression equation was then applied to the VEMAP II data to provide an estimate for Hubbard Brook – a simple linear calibration between the two data sets. Both data sets are monthly mean values for 1956 to 1993.
For radiation data, both VEMAP and Hubbard Brook report total global radiation in MJ/m2.day. As PnET uses PAR (umoles/m2.sec), we have converted from one to the other using
umoles = MJ X 2.0513 umole X day ------------------ ------------------ ------------------ ------------------ m2.sec m2.day MJ seconds of daylight
Note that PAR is expressed per second of daylight hours to be comparable with physiological measurements. This conversion requires a calculation of the number
of seconds between sunrise and sunset for each day. Radiation data at HB begin in 1960, so the regression period is 1960-1993.
| Acronyms: | |
| CASTNET: | Clean Air Status and Trends Network |
| NADP: | National Atmospheric Depostion Program |
| VEMAP: | Vegetation/Ecosystem Modeling and Analysis Project |
The equations resulting from the regressions between Hubbard Brook data and VEMAP II data for the period of measurement at Hubbard Brook are (expressed as HB data predicted from VEMAP II data):
| Tmax | = | 0.948319 | * | vemapII | - | 1.925 | (r2=0.99) | |
| Tmin | = | 0.970413 | * | vemapII | + | 0.571158 | (r2=0.99) | |
| Precip. | = | 0.9922 | * | vemapII | + | 2.078 | (r2=0.74) | |
| PAR | = | 0.7037 | * | vemapII | + | 88.37 | (r2=0.85) |
As you can see, the regressions for temperature were excellent, those for precipitation and radiation, less so. Slopes are close to 1 for all but PAR. Graphs of HB versus VEMAP II data follow.
 |  |
 |  |