Quantifying climatological precipitation and soil moisture as well as interannual variability and trends requires extensive observation. This work focuses on the analysis of available precipitation and soil moisture data and the development of new ways to estimate these quantities. Precipitation and soil moisture characteristics are highly dependent on the spatial and temporal scales. We begin at the point scale, examining hourly precipitation and soil moisture at individual gauges. First, we focus on the Walnut Gulch Experimental Watershed (WGEW), a 150 km² area in southern Arizona. The watershed has been measuring rainfall since 1956 with a very high density network of approximately 0.6 gauges per km². Additionally, there are 19 soil moisture probes at 5 cm depth with data starting in 2002. In order to extend the measurement period, we have developed a water balance model which estimates monsoon season (Jul-Sep) soil moisture using only precipitation for input, and calibrated so that the modeled soil moisture fits best with the soil moisture measured by each of the 19 probes from 2002-2012. This observationally constrained soil moisture is highly correlated with the collocated probes (R=0.88), and extends the measurement period from 10 to 56 years and the number of gauges from 19 to 88. Then, we focus on the spatiotemporal variability within the watershed and the ability to estimate area averaged quantities. Spatially averaged precipitation and observationally constrained soil moisture from the 88 gauges is then used to evaluate various gridded datasets. We find that gauge-based precipitation products perform best followed by reanalyses and then satellite-based products. Coupled Model Intercomparison Project Phase 5 (CMIP5) models perform the worst and overestimate cold season precipitation while offsetting the monsoon peak precipitation forward or backward by a month. Satellite-based soil moisture is the best followed by land data assimilation systems and reanalyses. We show that while WGEW is small compared to the grid size of many of the evaluated products, unlike scaling from point to area, the effect of scaling from smaller to larger area is small. Finally, we focus on global precipitation. Global monthly gauge based precipitation data has become widely available in recent years and is necessary for analyzing the climatological and anomaly precipitation fields as well as for calibrating and evaluating other gridded products such as satellite-based and modeled precipitation. However, frequency and intensity of precipitation are also important in the partitioning of water and energy fluxes. Therefore, because daily and sub-daily observed precipitation is limited to recent years, the number of raining days per month (N) is needed. We show that the only currently available long-term N product, developed by the Climate Research Unit (CRU), is deficient in certain areas, particularly where CRU gauge data is sparse. We then develop a new global 110-year N product, which shows significant improvement over CRU using three regional daily precipitation products with far more gauges than are used in CRU.
| Identifer | oai:union.ndltd.org:arizona.edu/oai:arizona.openrepository.com:10150/621434 |
| Date | January 2016 |
| Creators | Stillman, Susan, Stillman, Susan |
| Contributors | Zeng, Xubin, Zeng, Xubin, Dominguez, Francina, Niu, Guo-Yue, Troch, Peter |
| Publisher | The University of Arizona. |
| Source Sets | University of Arizona |
| Language | en_US |
| Detected Language | English |
| Type | text, Electronic Dissertation |
| Rights | Copyright © is held by the author. Digital access to this material is made possible by the University Libraries, University of Arizona. Further transmission, reproduction or presentation (such as public display or performance) of protected items is prohibited except with permission of the author. |
Page generated in 0.0019 seconds