Pollen-based climate reconstructions and syntheses in Europe
Fuente:
PANGAEA
Guardado en:
| Autores principales: | , |
|---|---|
| Formato: | Dataset Open Access |
| Lenguaje: | en |
| Publicado: |
PANGAEA
2025
|
| Materias: | |
| Acceso en línea: | |
| Etiquetas: |
Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
|
| _version_ | 1867168236277071872 |
|---|---|
| author | Geng, Rongwei Weinelt, Mara |
| author_facet | Geng, Rongwei Weinelt, Mara |
| collection | Datos científicos de ciencias marinas y ambientales |
| contents | A fossil pollen dataset distributed across Europe (10° W - 43° E, 33° - 71° N) comprising 520 records was extracted from the LegacyPollen 1.0 database (Herzschuh et al., 2022) to reconstruct climatic variables including Annual temperature (TANN), Annual precipitation (PANN), Winter Temperature (December, January, February; TDJF), Summer Temperature (June, July, August; TJJA). Short records not reaching beyond 1 ka BP were also excluded to keep the dataset refined, as the syntheses aim to cover the entire Holocene (i.e., 11-1 ka BP). The modern pollen training dataset was integrated from Legacy Climate 1.0 (Herzschuh et al., 2023) and the EMPD2 (Davis et al., 2020). Two different approaches were applied in parallel to reconstruct climate variables from fossil pollen assemblages, namely Modern Analogue Technique (MAT) and Weighted Averaging Partial Least Squares (WAPLS). Reconstruction uncertainties were provided as Root Mean Squared Errors of Prediction (RMSEPs). All the reconstructions and tests were conducted using the rioja and analogue packages in R (R Core Team, 2019). The synthesized results were interpolated from all reconstructed climate records. The mean value of reconstructed climatic variables with the same ages was calculated before any interpolations. Due to the different chronological resolution of the time series, the sequences were then interpolated to equidistant time series of 50-year intervals. Two different interpolation methods were applied in R. The first is to use the interp.dataset function from rioja package with loess regression to interpolate the dataset as a whole. The second is to interpolate each complete record that can cover the Holocene (i.e., 11-1 ka) and has a mean resolution of less than 1ka separately using the corit package with linear regression and then calculate the mean of these records. To perform the latter interpolation, a total of 214 records covering the entire period between 11-1 ka BP were used. The Root Mean Squared Errors (RMSEs) were calculated for the synthesis results. |
| format | Dataset Open Access |
| id | pangaea_https___doi_org_10_1594_PANGAEA_980704 |
| institution | PANGAEA |
| language | en |
| publishDate | 2025 |
| publisher | PANGAEA |
| record_format | pangaea |
| spellingShingle | Pollen-based climate reconstructions and syntheses in Europe Geng, Rongwei Weinelt, Mara Compilation; Europe; File content; Holocene; LegacyPollen_1.0_Europe; Office Open XML Workbook; Office Open XML Workbook (File Size); paleoclimatology; Pollen; Quantitative reconstruction A fossil pollen dataset distributed across Europe (10° W - 43° E, 33° - 71° N) comprising 520 records was extracted from the LegacyPollen 1.0 database (Herzschuh et al., 2022) to reconstruct climatic variables including Annual temperature (TANN), Annual precipitation (PANN), Winter Temperature (December, January, February; TDJF), Summer Temperature (June, July, August; TJJA). Short records not reaching beyond 1 ka BP were also excluded to keep the dataset refined, as the syntheses aim to cover the entire Holocene (i.e., 11-1 ka BP). The modern pollen training dataset was integrated from Legacy Climate 1.0 (Herzschuh et al., 2023) and the EMPD2 (Davis et al., 2020). Two different approaches were applied in parallel to reconstruct climate variables from fossil pollen assemblages, namely Modern Analogue Technique (MAT) and Weighted Averaging Partial Least Squares (WAPLS). Reconstruction uncertainties were provided as Root Mean Squared Errors of Prediction (RMSEPs). All the reconstructions and tests were conducted using the rioja and analogue packages in R (R Core Team, 2019). The synthesized results were interpolated from all reconstructed climate records. The mean value of reconstructed climatic variables with the same ages was calculated before any interpolations. Due to the different chronological resolution of the time series, the sequences were then interpolated to equidistant time series of 50-year intervals. Two different interpolation methods were applied in R. The first is to use the interp.dataset function from rioja package with loess regression to interpolate the dataset as a whole. The second is to interpolate each complete record that can cover the Holocene (i.e., 11-1 ka) and has a mean resolution of less than 1ka separately using the corit package with linear regression and then calculate the mean of these records. To perform the latter interpolation, a total of 214 records covering the entire period between 11-1 ka BP were used. The Root Mean Squared Errors (RMSEs) were calculated for the synthesis results. |
| title | Pollen-based climate reconstructions and syntheses in Europe |
| topic | Compilation; Europe; File content; Holocene; LegacyPollen_1.0_Europe; Office Open XML Workbook; Office Open XML Workbook (File Size); paleoclimatology; Pollen; Quantitative reconstruction |
| url | https://doi.org/10.1594/PANGAEA.980704 |