Fine-Scale Climate Mapping
From its inception, the CLIMAS research agenda has been guided by feedback from a wide variety of stakeholders who use climate information to make decisions about their operations. Many of the people interviewed in the CLIMAS pilot study expressed dissatisfaction with the way climate information systems portray rural climate conditions.
There is a sense of metropolitan bias in the information they receive and rely upon, particularly for those located some distance from the metropolitan centers of Tucson or Phoenix. This bias neglects the very different climates of regions outside of Tucson or Phoenix. Moreover, stakeholders consistently expressed an interest in easily accessible climate data at "local" spatial scales, e.g., the size of small ranching and farming operations.
In response to stakeholder requests for fine-scale climate data, we created a research project to develop models for interpolating winter climate data for the Southwest to 1 km2 resolution. Our goal was to make this climate knowledge useful to regional stakeholders of climate information as well as other researchers.
The winter season was selected as the initial temporal study period, because of the crucial role precipitation plays in the recharge of dams, aquifers, and reservoirs, and for overlap with the CLIMAS paleoclimate project that deals with winter moisture. Topography is the major source of spatial variability in climate data at these spatial scales, and therefore the models were based on terrain variables such as elevation, slope, aspect, latitude, and longitude.
This project aimed to identify patterns of local and regional climate variability in these datasets and the atmospheric features that control southwestern climate. Results were intended to be suitable for use in climatic and other environmental studies by public and private stakeholders.
Background
Dry and hot! For many people, these two words sum up the climate of the Southwestern United States, but the climate of the Southwest is much more complex. While low deserts of the Southwest experience searing heat and desiccating winds in the early summer, its forested mountains and plateaus endure biting cold and drifting snow in the heart of winter. The Southwest may be drenched by torrential monsoon thunderstorms in July and August, yet it can warm gently under fair skies from fall to spring.
Rainfall in the Southwest is especially variable, with regional floods or droughts severe enough to affect both indigenous and modern civilizations on time scales ranging from single growing seasons to multiple years, even decades.
Interannual and decadal precipitation variability plays major societal and ecological roles in western North America, for example in areas such as ranching, farming, tourism, urban water management, and regional power production within the Southwest. Moreover, not all parts of the Southwest are affected to the same degree, or even at the same time, by weather events and long-term climate patterns. During some summers rainfall might be abundant in southwestern New Mexico, but scanty in northwestern Arizona. Long-term droughts, such as the one that gripped the region during the 1950s, might be far more severe in one part of the 2-state region—as evidenced by the profound effect of the 1950s drought on the forests of eastern New Mexico.
Thus, data were needed by researchers to investigate climate and weather at a variety of time scales and over spatial scales ranging from tens to thousands of square kilometers. Individuals and agencies that use climate information to make decisions also need data that allow them to look at past climate variations that affect businesses and land management operations ranging from a few square kilometers to all of Arizona and New Mexico.
Currently, free climate data for Arizona and New Mexico consist of records for individual meteorological stations, regional records for large climate divisions (some of which are as large as several counties) and coarse-scale gridded records developed from interpolation of station records, without regard to elevation—which may vary dramatically over very short distances.
Research Methods
Our goal was to develop a methodology that could ultimately be used to produce several datasets for the region, and which was sufficiently flexible that we could apply it to various timescales in both the instrumental and paleoclimatic records.
It was important for us to account for the large topographic variations across the Southwest. Thus we developed a technique, which uses topographic data, including elevation, slope and aspect, to estimate winter (December–March) temperature and precipitation throughout Arizona and New Mexico at 1 x 1 km (0.6 x 0.6 mi.) resolution. Data from 662 temperature stations and 572 precipitation stations were used to create the interpolated data. Exploratory analyses 
showed that a single regression model was sufficient for creating gridded temperature datasets at 1 x 1 km resolution. For precipitation, a series of sub-regional models was used rather than a region-wide model (Brown and Comrie, 2002).
The final temperature (Figure 1) and precipitation (Figure 2) models explain 98 percent and 63 percent of the variance in the station data, respectively. Climate anomalies were calculated by examining the differences between each temperature and precipitation station and its respective modeled values. These anomaly data were used to produce winter temperature and precipitation maps at 1 x 1 km resolution for the period 1961 to 1999.
The final portion of the study is an investigation of the connections between winter climate in Arizona and New Mexico and large-scale climate patterns in the Pacific Ocean.
We have compared the 1 x 1 km resolution winter climate data with well-documented climatic indices representing patterns of atmospheric and oceanic variability in the North and East Pacific Ocean for the period 1961 to 1999. We calculated linear statistics to quantify the relationships between our fine-scale winter climate and the climate indices.
Related Publications
Brown, D. and A. Comrie. 2004. A winter precipitation 'dipole' in the western United States associated with multidecadal ENSO variability. Geophysical Research Letters, 31, doi:10.1029/2003GL018726.
Brown, D. and A. Comrie. 2002. Spatial modeling of winter temperature and precipitation in Arizona and New Mexico, U.S.A. Climate Research, 22:115-128.
Brown, D. and A. Comrie. 2002. Sub-regional seasonal precipitation linkages to SOI and PDO in the southwest United States. Atmospheric Science Letters, 3:94-102.
Daly, C., R. Neilson, and D. Phillips. 1994. A statistical-topographic model for mapping climatological precipitation over mountainous terrain. Journal of Applied Meteorology, 33:140-158.