U42B-01
The North American Drought Atlas: Tree-Ring Reconstructions of Drought Variability for Climate Modeling and Assessment
The North American Drought Atlas describes a detailed reconstruction of drought variability from tree rings over most of North America for the past 500-1000 years. The first version of it, produced over three years ago, was based on a network of 835 tree-ring chronologies and a 286-point grid of instrumental Palmer Drought Severity Indices (PDSI). These gridded PDSI reconstructions have been used in numerous published studies now that range from modeling fire in the American West, to the impact of drought on palaeo-Indian societies, and to the determination of the primary causes of drought over North America through climate modeling experiments. Some examples of these applications will be described to illustrate the scientific value of these large-scale reconstructions of drought. Since the development and free public release of Version 1 of the North American Drought Atlas (see http:iridl.ldeo.columbia.edu/SOURCES/.LDEO/.TRL/.NADA2004/.pdsi-atlas.html), great improvements have been made in the critical tree-ring network used to reconstruct PDSI at each grid point. This network has now been enlarged to 1743 annual tree-ring chronologies, which greatly improves the density of tree-ring records in certain parts of the grid, especially in Canada and Mexico. In addition, the number of tree-ring records that extend back before AD 1400 has been substantially increased. These developments justify the creation of Version 2 of the North American Drought Atlas. In this talk I will describe this new version of the drought atlas and some of its properties that make it a significant improvement over the previous version. The new product provides enhanced resolution of the spatial and temporal variability of prolonged drought such as the late 16th century event that impacted regions of both Mexico and the United States. I will also argue for the North American Drought Atlas being used as a template for the development of large-scale drought reconstructions in other land areas of the Northern Hemisphere where sufficient tree-ring data exist. By doing so, the importance of this product to the modeling community will be significantly enhanced.
U42B-02
Preliminary spatial Standardized Precipitation Index (SPI) drought reconstructions for central and northern Mexico
We developed a spatial drought reconstruction of the Standardized Precipitation Index (SPI) for northern and central Mexico based on the relationship between tree growth and climate. Many previous tree-rings studies have successfully reconstructed drought indices as Palmer Drought Severity Index (PDSI), although SPI has become a widely accepted drought index; to our knowledge no SPI reconstructions have been developed using tree-ring chronologies over our study region. The methodology used in this work was the Point by Point Regression (PPR). The data set used was a network of tree-ring chronologies for Mexico and south United States and a 7-month Standardized Precipitation Index (SPI) for the season November - May. The observed SPI was based on the Pearson Type III distribution (Guttman 1999), calculated from a 1-degree monthly precipitation dataset (Chen et al 2002). In order to maximize the number of tree-ring chronologies we set our calibration and reconstruction periods for 1948-1985 and 1801- 1985, respectively. According with our analysis, El Nino events are mainly associated with wetter condition over northern Mexico, while La Nina events are characterized by dry conditions. This reconstruction reproduces the 1950s drought which affected most of northern Mexico, the Great Plains and the southwestern United States. Results are compared with other reconstructed gridded drought indices available for the same area (Cook et al 2004).
U42B-03
Relationships between tree growth in western Canada and the Pacific Ocean
Tree-ring chronologies have been developed from a number of species growing in different environments (e.g. upper and lower treeline) in the western Cordillera of Canada (from the U.S. border to the Yukon; and from interior B.C. across the Rockies). Chronologies from these different species have a range of responses to climate parameters and have been useful for reconstructing precipitation and temperature. More recently these chronologies have been combined to develop reconstructions of climate-related parameters including glacier mass balance and streamflow both of which are strongly related to winter precipitation receipts. In this paper we present results from work that explores relationships between measured and reconstructed climate variables from the western Cordillera and conditions in the Pacific Ocean. We focus on the extent to which variations in the Pacific Ocean (captured by large-scale indices and correlation patterns with SSTs) can explain variance in climate parameters from the region for different seasons (e.g. summer drought versus winter precipitation). We also comment on the utility of these tree-ring data for reconstructing past conditions in the Pacific based on the strength of observed relationships. Preliminary results from recent sampling expeditions that attempt to target winter snowpack sensitive tree-ring chronologies are also presented.
U42B-04
Medieval Drought in the Upper Colorado River Basin
Paleoclimatic records have consistently identified the Medieval Climate Anomaly (MCA), approximately A.D. 900- 1300, as unusual for the incidence of multi-decadal drought in the western United States. Four newly developed tree-ring chronologies derived from living trees, standing dead trees and logs are examined for evidence of a MCA drought signature in the Upper Colorado River Basin (UCRB). The four chronologies extend to before A.D. 1000, and each has a statistically significant significant linear relationship with annual flow of the Colorado River from 1906 to present (23-48 percent of the flow variance explained). A flow reconstruction based on three chronologies with acceptably high sample depth in the MCA begins in A.D. 762, accounts for 60 percent of the flow variance in the 1906-2003 calibration period, and identifies a drought in the mid-1100s as the most severe multi-decadal drought in the long-term record. Highlights of this inferred drought are 13 consecutive years of below normal flow (A. D. 1143-1155) and a broad window of 62 years (A. D. 1118-1179) with no years of much-above-normal flow. The individual chronologies reveal a spatial pattern of heightened intensity of drought toward the western part of the UCRB, although the drought was also present in the Colorado Rockies. Comparison with other regional paleoclimatic reconstructions from the western United States shows the results are consistent with dry, and probably warm, conditions over the intermountain region. Differences are noted in timing of the mid-1100s drought in the UCRB and epic MCA droughts in the Sierra Nevada of California inferred from tree-ring data and other paleoclimatic evidence. The results emphasize the importance of striving for greater spatial resolution in the paleoclimatic record of drought in western North America.
U42B-05
Multidisciplinary Investigations of Climate Change in Mexico: Reports from a Collaborative Academic Network
Unusually severe or prolonged drought ranks among the most devastating of climate changes in Mexico, and has
contributed to wildfires, crop failure, livestock death and famine across the country throughout its history and pre-
history (Liverman, 1999; Florescano, 1980). Some of these droughts are thought to be associated with the El
Niño Southern Oscillation (ENSO) - the most prominent signal in inter-annual climate variability in Mexico and
the subject of increasing interest following the effects of the severe El Niño of 1997-98 (Magaña et al.,
2003). Obtaining detailed records of droughts and gaining a better understanding of how rainfall variability has
affected Mexico's populations in the past, and is affecting society today, are essential steps toward designing
appropriate drought preparedness, mitigation and relief programmes both in Mexico and other vulnerable regions
of the world (ISDR, 2002).
This paper reports on the development of Collaborative Academic Network entitled "Multi-disciplinary
investigations of climate change in Mexico". The network represents an important research initiative, funded by
the Leverhulme Trust, involving individuals from Mexico, the United Kingdom and the United States with
established reputations for research in complementary methodological approaches to understanding climate
change and drought in Mexico. The aim of the network is to draw on internationally recognised expertise in
dendroclimatology, palaeolimnology, climate history and modern climatology, with the following objectives:
• to improve understanding of the climatic mechanisms that cause drought across Mexico and how these vary
spatially;
• to identify past, present and future impacts of, and responses to, drought in Mexico;
• to develop a more comprehensive understanding of climate change and its impacts across Mexico on a range
of time scales.
Here I report on the progress of the network so far in its aims to facilitate the integration and development of
research at the forefront of Mexican climate change studies.
http:www.nottingham.ac.uk/geography/network-mexico/
U42B-06
Pacific Ocean Forcing and the Development of Prolonged ‘Perfect Droughts' in Southern California and Adjacent Western North America Over the Holocene
A ‘perfect drought' in terms of southern California is defined here as a prolonged drought that impacts southern California and its major extra-regional water sources in northern California and the upper Colorado Basin. Instrumental climate records over the past century demonstrate that such super-regional droughts have occurred regularly, but generally persist less than 5 years. A notable example is the prolonged drought in the late 1980's - early 1990's which severely impacted southern California and areas northeastward to the midwestern United States and adjacent portions of the Canadian prairies. The core period of this drought from 1988 through 1990 was associated with cool conditions in the eastern tropical Pacific, and 40 to 50 degrees N latitude across the northern Pacific. During that time a persistent positive 700 mb height anomaly was anchored over the same region of the northeastern Pacific. Dendrohydrological estimates of the Palmer Drought Severity Index (PDSI) for southern California and annual discharge for the Sacramento and Colorado river systems for the past 1000 years indicate generally dry conditions during the Medieval period with two particularly notable periods of widespread ‘perfect drought' in the 11th and 12th centuries. Prolonged dry conditions in the west during this time can be linked through both empirical evidence and climate model experiments to the establishment of persistent cool sea surface temperatures (SST's) in the eastern tropical Pacific and northeastern Pacific. A Holocene length record of aridity obtained from lake sediments in the Sierra Nevada shows that such prolonged episodes of aridity have occurred prior to the last millennium and that prolonged aridity on the scale of centuries to millennia can directly be linked to variations in eastern Pacific SST's as resolved by paleoceanographic records.
U42B-07
Holocene hydroclimate of the Sonoran desert: Results from Cave of the Bells, Arizona
In the desert of North America, water availability fundamentally limits the scope of human and ecological activity. Understanding the natural variability of the region's hydroclimate and its sensitivity to radiative forcings is important for anticipating future changes in water availability. Here we present an isotopic record from a speleothem collected in southeastern Arizona that spans 3.5-6.9 ky BP at interannual resolution. Measurements on modern cave and meteoric waters suggest that the cave currently receives only winter moisture, despite the fact that half the region's precipitation falls during the summer monsoon. Speleothem isotopic values are higher than present by about 3 per mil, a change too large to be explained by warming/drying. We suggest this change is due to an increase in the amount of summer monsoon precipitation reaching the cave, consistent with orbital forcing of a stronger monsoon. The Holocene record of climate in the southwest US is complex, but many records support an intensified monsoon, and some suggest drying that is confined to winter. Our record is consistent with an increase of summer moisture relative to winter during the mid-Holocene, a scenario that explains a wide range of divergent paleoclimatic results. The record also shows substantial multidecade-century scale variability. Spectral analysis of discrete intervals reveals peaks in the early part of the record near 233 and 142 years, giving way to a broad concentration of multidecadal variance at periods between 50-70 years since 4.9 ky BP. The recent transition to higher frequency variability is consistent with the hypothesized evolution of ENSO towards more frequent occurrence. Our record clearly illustrates both the importance of multidecade-century time scales of hydroclimate variability and the sensitivity of the North American monsoon to radiative forcing. We hope to quantify monsoon changes by expanding our work into northern Mexico, where the summer monsoon dominates annual rainfall.
U42B-08
Tree Ring Chronologies in Mexico and Dendroclimatic and Ecological Applications
The understanding of historic hydroclimate variability is basic to determine the impact of atmospheric circulatory patterns and to plan for a proper management of limited water resources and ecosystem conservation purposes. The objective of this study was to develop a network of tree-ring chronologies for climate reconstruction and to analyze the influence of the ENSO warm phase in northern Mexico by using the Tropical Rainfall Index. Climatic sensitive tree-ring chronologies were developed in mountain ranges and riparian ecosystems of the Sierras Madre Oriental and Occidental, and central Mexico. A grid of over 30 new Douglas-fir, baldcypress, and pinyon pine chronologies were developed and seasonal winter-spring and summer precipitation reconstructions have been produced for northern and central Mexico. The seasonal winter-spring precipitation reconstructions extended 353 years (1450 - 2002) and 530 years (1472 - 2002) for Chihuahua, 228 years (1765 - 1993) and 504 years (1500 - 2003) for Durango, 602 years (1400 - 2002) for Nuevo Leon, 522 years (1474 - 1995) for Tamaulipas; and 342 years (1659 - 2001) and 410 years (1595 - 2004) for Coahuila. Some of the low frequency events were specific for each reconstruction, but common low frequency events (decadal resolution) were present in most of the reconstructions; specific cases are the droughts of the 1780s, 1810s, 1860s, 1870s, 1950s, and 1990s; and the wet episodes of the 1820s, 1830s, and 1890s.Trends in dry or wet periods were disrupted by above or below normal precipitation as affected by the ENSO phenomena, especially in the winter- spring period when this circulatory pattern produced in times abundant rains in northern Mexico. However, the ENSO influence on winter-spring precipitation has fluctuated through time. Cold fronts and hurricanes impacting the Gulf of Mexico may explain some of the hydrological behavior detected for northeastern Mexico. These results indicate that winter-spring hydroclimate variability in northern Mexico is influenced by a range of atmospheric circulatory patterns, and a greater grid of tree-ring chronologies should be developed to better explain climatic variability in this region. Currently, a couple of summer precipitation reconstructions have been developed and others are in process, but future research will focus on this issue. Summer rainfall represents over 70% of the annual precipitation in Mexico and provides water for agriculture, forest productivity, and other uses. On the other hand, fire frequency and forest dynamic studies are in process for restoration and conservation purposes.