Informations et ressources scientifiques
sur le développement des zones arides et semi-arides

Accueil du site → Master → Etats Unis → 2013 → Improving the Palmer drought severity index by incorporating snow and frozen ground

UNIVERSITY OF NORTH DAKOTA (2013)

Improving the Palmer drought severity index by incorporating snow and frozen ground

Qiu, Shaoyue

Titre : Improving the Palmer drought severity index by incorporating snow and frozen ground

Auteur : Qiu, Shaoyue

Université de soutenance : UNIVERSITY OF NORTH DAKOTA

Grade : Master of Science (MS) 2013

Résumé
Drought causes extensive damage and affects a significant number of people. To quantify the severity of drought and to better monitor drought, drought indices are necessary. The Palmer Drought Severity Index (PDSI) has its advantages of comprehensive and taking the characteristics of drought into concern so that it was regarded as the milestone in the revolution of drought indices (Heim, 2002). The PDSI has been widely used, tested, and modified since its development in 1965. However, a commonly documented limitation to this index is the lack of consideration for snow and the treatment of ground status. In this study, a simplified snow model is included in the self-calibrated PDSI model, which is run in a monthly time step over the continental United States (CONUS) for the past 32 years (1979–2010). In the modified PDSI model, the form of precipitation is based on an air temperature threshold and moisture has been withheld and redistributed into the system based on the accumulation and melt of snow. With the snow processes included, all the model variables (evapotranspiration, soil moisture recharge/loss, surface runoff) decreased in winter and increased in spring melt time and soil moisture remained the same in winter and got recharged in spring, the difference between the soil moisture in the original PDSI model without snow effect and the soil moisture in the PDSI model with snow effect is small. The climatically appropriate precipitation increased more than 200% in northern latitude and mountain regions in spring melt seasons. The absolute value of moisture departure also decreased in winter and increased in summer, which means the inclusion of snow processes made the moisture departure increase during wet condition and decrease for dry condition (varies more). Inclusion of the snow model also allows the PDSI to better capture spring flooding events, which are caused by snowmelt ; monitoring drought events also has been improved due to the changes in duration factor for the modified model with snow processes. Finally, the general moisture conditions as well as the trend of moisture change have been examined using both the original model and the modified model including snow. Both models show similar characteristics over the CONUS for the past three decades. The inclusion of the snow model does not qualitatively change these results, and has little effect on the spatial comparability of the index. Effect of frozen soil has also been examined. In this study, this effect is simply tested by shutting down infiltration when ground is frozen. Due to the time lag between ground status and air temperature (which is the determinant for the calculated snowmelt rate), and also the different spatial resolution for these two data sets, most of the spring snowmelt in western US, especially the northern Rocky Mountain regions didn’t infiltrate the soil system and became surface runoff. This loss of moisture caused the soil moisture in those regions decreased from a climate mean 35cm to only 6 cm, which further made the longitude-averaged annual mean PDSI decreased 0.5 to 0.7 index value.

Search Oxford Libraries Online

ProQuest Dissertations and Theses

Page publiée le 20 décembre 2014, mise à jour le 6 mars 2019