An investigation into wheat's vulnerability in the western U.S.
Presentation Type
Oral Presentation
Abstract/Artist Statement
As both world population grows and diets change, global food demand is expected to double by 2050. However, as regional climates shift, changes in average weather conditions will likely have significant consequences for croplands. Thus constraining the impact the climate has on croplands is an essential task of the 21st century.
Heat and drought metrics have already been correlated with county level wheat yield data and suggest vulnerabilities to both high temperatures and lack of rainfall. Mechanistic models also suggest vulnerabilities to lack of moisture and high temperatures. Most of the empirical studies utilize county or larger scaled data. However, field scale yield data is becoming more popular because of potential spatial heterogeneity across county scales. A likely trade off is that it may be harder to extrapolate to larger (i.e. country, continental, or global) scales. Here, we utilize a remote sensing metric, Normalized Difference Vegetation Index(in lieu of yield data), which quantifies how much green light the satellite receives. We use Landsat’s 30m resolution, thereby keeping similar resolution to field scale studies but being able to sample anywhere across the globe. We hypothesize that wheat in the 21st century grown in the western U.S. is primarily constrained by lack of water, both the supply and demand. And as such, we utilize metrics such as vapor pressure deficit and cumulative dry days to pull out more nuanced effects weather may have on wheat’s greenness. Furthermore, we hypothesize that not all wheat growing regions will have the same sensitivity to changes in meteorological variables. Thus, we utilize a self organizing map and k means approach to cluster agro-climatological regions and compare their relative meteorological sensitives. We find that North Eastern North Dakota is rather resilient to changes in water supply and demand and that Eastern Montana is much more vulnerable. We also find that areas with lots of irrigation (i.e. Western WA, South East Idaho) are somewhat buffered from drought, but by no means completely. These findings can offer nuance to discussions on farmer adaptation, water resource planning, and help further constrain mechanistic crop models.
Mentor Name
Marco Maneta
An investigation into wheat's vulnerability in the western U.S.
UC 330
As both world population grows and diets change, global food demand is expected to double by 2050. However, as regional climates shift, changes in average weather conditions will likely have significant consequences for croplands. Thus constraining the impact the climate has on croplands is an essential task of the 21st century.
Heat and drought metrics have already been correlated with county level wheat yield data and suggest vulnerabilities to both high temperatures and lack of rainfall. Mechanistic models also suggest vulnerabilities to lack of moisture and high temperatures. Most of the empirical studies utilize county or larger scaled data. However, field scale yield data is becoming more popular because of potential spatial heterogeneity across county scales. A likely trade off is that it may be harder to extrapolate to larger (i.e. country, continental, or global) scales. Here, we utilize a remote sensing metric, Normalized Difference Vegetation Index(in lieu of yield data), which quantifies how much green light the satellite receives. We use Landsat’s 30m resolution, thereby keeping similar resolution to field scale studies but being able to sample anywhere across the globe. We hypothesize that wheat in the 21st century grown in the western U.S. is primarily constrained by lack of water, both the supply and demand. And as such, we utilize metrics such as vapor pressure deficit and cumulative dry days to pull out more nuanced effects weather may have on wheat’s greenness. Furthermore, we hypothesize that not all wheat growing regions will have the same sensitivity to changes in meteorological variables. Thus, we utilize a self organizing map and k means approach to cluster agro-climatological regions and compare their relative meteorological sensitives. We find that North Eastern North Dakota is rather resilient to changes in water supply and demand and that Eastern Montana is much more vulnerable. We also find that areas with lots of irrigation (i.e. Western WA, South East Idaho) are somewhat buffered from drought, but by no means completely. These findings can offer nuance to discussions on farmer adaptation, water resource planning, and help further constrain mechanistic crop models.