VEMAP Phase 2 bioclimatic database. I. Gridded historical (20th century) climate for modeling ecosystem dynamics across the conterminous USA

T. G F Kittel*, N. A. Rosenbloom, J. A. Royle, C. Daly, W. P. Gibson, H. H. Fisher, P. Thornton, D. N. Yates, S. Aulenbach, C. Kaufman, R. McKeown, D. Bachelet, D. S. Schimel, R. Neilson, J. Lenihan, R. Drapek, D. S. Ojima, W. J. Parton, J. M. Melillo, D. W. KicklighterH. Tian, A. D. McGuire, M. T. Sykes, B. Smith, S. Cowling, T. Hickler, I. C. Prentice, S. Running, K. A. Hibbard, W. M. Post, A. W. King, T. Smith, B. Rizzo, F. I. Woodward

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

39 Citations (Scopus)

Abstract

Analysis and simulation of biospheric responses to historical forcing require surface climate data that capture those aspects of climate that control ecological processes, including key spatial gradients and modes of temporal variability. We developed a multivariate, gridded historical climate dataset for the conterminous USA as a common input database for the Vegetation/Ecosystem Modeling and Analysis Project (VEMAP), a biogeochemical and dynamic vegetation model intercomparison. The dataset covers the period 1895-1993 on a 0.5° latitude/longitude grid. Climate is represented at both monthly and daily timesteps. Variables are: precipitation, mininimum and maximum temperature, total incident solar radiation, daylight-period irradiance, vapor pressure, and daylight-period relative humidity. The dataset was derived from US Historical Climate Network (HCN), cooperative network, and snowpack telemetry (SNOTEL) monthly precipitation and mean minimum and maximum temperature station data. We employed techniques that rely on geostatistical and physical relationships to create the temporally and spatially complete dataset. We developed a local kriging prediction model to infill discontinuous and limited-length station records based on spatial autocorrelation structure of climate anomalies. A spatial interpolation model (PRISM) that accounts for physiographic controls was used to grid the infilled monthly station data. We implemented a stochastic weather generator (modified WGEN) to disaggregate the gridded monthly series to dailies. Radiation and humidity variables were estimated from the dailies using a physically-based empirical surface climate model (MTCLIM3). Derived datasets include a 100 yr model spin-up climate and a historical Palmer Drought Severity Index (PDSI) dataset. The VEMAP dataset exhibits statistically significant trends in temperature, precipitation, solar radiation, vapor pressure, and PDSI for US National Assessment regions. The historical climate and companion datasets are available online at data archive centers.

Original languageEnglish
Pages (from-to)151-170
Number of pages20
JournalClimate Research
Volume27
Issue number2
Publication statusPublished - 7 Oct 2004
Externally publishedYes

Keywords

  • Climate change
  • Climate dataset
  • Climate variability
  • Ecological modeling
  • Ecosystem dynamics
  • Geostatistics
  • Palmer Drought Severity Index
  • PDSI
  • Vegetation
  • VEMAP

Fingerprint Dive into the research topics of 'VEMAP Phase 2 bioclimatic database. I. Gridded historical (20th century) climate for modeling ecosystem dynamics across the conterminous USA'. Together they form a unique fingerprint.

Cite this