Access to infrastructure is considered to influence the feasibility and efficacy of aid distribution programs in response to disasters and used to represent physical capital.Given that better access to power services may reduce the impacts of winter storms by providing alternative or additional assistance, access to facilities was used to represent physical capital.GIS data on power plants and facilities were obtained from the U.S.Environmental Protection Agency’s Facility Registry Service and Iowa Facility Explorer.The interviewed farmers also reported that a major winter storm loss on farms was from animal death caused by inadequate feed.Thus, feed supply was also considered as a physical capital indicator and represented by the 2012 feed expenditure data collected from USDA QuickStats.Human capital.Labor is considered to make a positive impact on vulnerability reduction because more family members can increase work efficiency during both events and subsequent recovery.This study used household size and labor expense as human capital indicators to represent the availability of labor engaged in adaptation.Education level, which is considered to increase the adaptive capacity by enhancing access to information , was also included to estimate human capital.The more skills and knowledge acquired, the more capability households have for emergency planning, recovery, and decision-making.Data on household size, labor expense, and education level were collected from the US Census Bureau.Social capital.Social organizations can improve adaptive capacity by enhancing social networking.Households with a membership to farm-related organizations are more likely to receive support or benefit from the professionals.To obtain information on membership with the agricultural organizations, a request was submitted to the contact on the Practical Farmer of Iowa website.Interview results also reveal that the reduction of storm losses can be attributed to the registration of insurance packages and government programs.More investment in government programs could provide more support during the storm recovery process.
The government program expense used in this study was retrieved from USDA QuickStats.Overall,mobile vertical farm a total of 12 adaptivity variables, 2 sensitivity variables, and 2 exposure variables were selected for the assessment of rural winter storm vulnerability.Socioeconomic statistics and spatial information were all aggregated to the census county level and standardized to Z-scores in SPSS before further analysis.There are 29 out of 60 significantly correlated pairs with a p-value of less than 0.050, indicating strong interrelationships between indicators.Hence, these indicators are considered suitable for factor analysis to extract principal components accounted for by the variable correlations.The correlation coefficients range from − 0.459 for farm income and natural shelter to 0.788 for farm income and labor expense.Counties planting more trees appear to receive lower income.Labor can increase farming productivity and, at the same time, require more investment, leading to the strongly positive relationship between farm income and labor expense.There is also a strong correlation between membership counts and education, indicating that counties with higher education levels are more likely to subscribe to farming associations.Among the selected 12 variables, poverty, energy, internet operations, and household size yielded low community values , suggesting that they would be weakly reflected via the extracted factors and thus be removed from factor analysis.Finally, with the remaining 8 variables, factor analysis extracted the first 3 factors that could yield a total of 85.124% of total variance explained , with an acceptable KMO value of 0.627.The Bartlett’s Test was statistically significant, indicating the high independency among the 8 variables.The loadings matrix in Table 5 shows the correlations of each variable with the three extracted components.Those with loadings greater than 0.800 are considered as salient indicators representing the three underlying dimensions of adaptive capacity determinants.The first factor is interpreted as farming economic status based on its salient indicators of labor expense, farming facilities, and farm income.This factor is considered to project adaptive capacity more accurately as it accounts for the largest total variance of the input variables.Economic conditions may be the most important determinant of adaptive capacity, probably because economic resources can facilitate technology implementation, ensure training opportunities, and lead to political influence.The second factor has high loadings on natural shelter and government programs, hence it is explained as environmental institutional capital.This factor may suggest a strong correlation between institutional efforts and the enhancement of environmental services.
For example, through general or continuous funding, the state of Iowa has a variety of conservation programs aimed to provide cost-sharing for tree planting on a highly erodible row crop and pasture land , potentially increasing farmers’ adaptive capacity to winter storms.The third component is highly correlated with education and organization membership.These indicators representing human capital and social capital are considered to affect innovative performance.Therefore, innovative capital is reasoned as the theme for the third component of adaptive capacity.The overall exposure rates are high in Northwest and Southeast Iowa due to high event frequency.This is consistent with the long history of severe winter storms and blizzards recorded for these regions.In contrast, eastern Iowa shows the lowest exposure scores.Sensitivity indicator scores were calculated by summing the standardized variable scores for animal sale and building age.As shown in Fig.4, counties peripheral to central Iowa tend to be more sensitive due to a high percentage of the total sale from animal commodities.From East to Central Iowa, the counties are light-colored, indicating low rates for building age and animal sale.This contributes to the notably least overall sensitivity for Polk County and its surrounding counties.Several counties score high in animal sale and/or building age, leading to their high overall sensitivity scores.Fig.5 shows the overall adaptive capacity and individual factor scores.Figure 5a shows that the adaptive capacity is low in most northwestern counties in Iowa and high in central Iowa and northeastern margins.It is noted from Fig.5b that counties in northern Iowa have higher rates for farming economic status as they have higher labor expense, farm-related income, and farming facilities than counties in the southernmost part of Iowa.Sioux appears to have the best farming economic status, as opposed to the metropolitan regions where farming-related investments are low.Fig.5c shows that the northwestern quarter of Iowa is low in environmental institutional capital, with limited natural shelter and low expense on government programs.This may be because the long-standing large tracts of wetlands concentrated in the northwest and north-central parts of Iowa have provided rich farmland for growing intensive crops.The increase of mono-cultures and the decrease in livestock pastures in the northwest could lead to the destruction of windbreaks.The patchwork of small, diversified fields that once were common remains in southeastern Iowa.
In northeastern Iowa, the rugged landscape with more wooded areas may have prevented farms from expanding to large industrialized operations, resulting in high index scores for environmental institutional capital.Fig.5d shows a concentration of innovative capital in the metropolitan areas of central Iowa and cold spots in northwestern and southeastern Iowa.Fig.6 illustrates the overall vulnerability for all Iowa counties calculated using the overall exposure, sensitivity, and adaptive capacity scores.In general, southern counties such as Adams and Union are remarkably vulnerable to winter storms, perhaps because much of their land areas in southern Iowa is used for perennial pastures , increasing their sensitivity.Highly vulnerable counties are also clustered in the Northwest where winter storm events are more frequent and in the Southeast where winter temperature deviation is higher, both reflecting high exposure.The vulnerability is low in central Iowa due to low sensitivity from East to Central Iowa, in particular in Polk and its adjacent metropolitan areas.Counties with low vulnerability are also found in northeastern Iowa where adaptive capacity is higher.Among different disaster types, winter storms receive limited attention, while they cause non-negligible costs.In Iowa, there appears a generally increasing trend in experiencing winter storm events, indicated by more above-average event occurrences in the recent past.Evaluating the vulnerability of farming communities to winter storms in Iowa has implications for identifying counties’ agricultural production prone to winter storms and thus reducing farm loss during winter storms by managing the vulnerability components, namely, exposure, sensitivity, and adaptive capacity.Exposure can be influenced by the increased population and assets at risk as a result of population growth in locations at risk from natural hazards , and storm impacts are likely to be worse in more populous areas than others.However, Polk County – the most populous county in Iowa – rated the least vulnerable to winter storms,vertical farming racks whilst it has relatively high exposure.Its low score in vulnerability may be due to their industry-oriented development that is more resistant to winter storms than farming activities.This indicates the severity of weather events is not necessarily consistent with the population pattern alone as it may vary depending on the specific disasters or economic structure.To explore the issue further, the difference between vulnerability level and factual on-farm loss in 2012 per county was calculated and illustrated in Fig.8.After scaling to the range of 0–1, the overall difference ranged from 0.009 for Johnson County to 0.88 for Van Buren County.
Counties graphed in the left half of Fig.8 show almost identical distributions of farm loss and vulnerability.This implies the selected indicators for winter storm vulnerability in the current study may be used to effectively evaluate the general farm losses for these counties for a given year.It is found the metropolitan county of Story has non-negligible farm loss and underpredicted vulnerability.This suggests the limitation in the current model that is unable to capture all critical factors to determine the area’s general farm loss.For example, farming intensity may scale the loss but is not considered in the model.Agricultural production characteristics such as the quantity of products vulnerable to other storm events as well as meteorological variability such as winter storm occurrence may also contribute to the discrepancy between empirical farm losses and predictions.To account for all counties’ general loss characteristics determined by factors not included in the current winter storm vulnerability model, the 2002-2017-census-year average farm loss was calculated.Several counties in the left half of Fig.8 show small differences between farm loss in 2012 and average farm loss, indicating these counties have relatively stable farm loss patterns and the current model can be used to evaluate their long-term general farm losses.On the other hand, counties displayed on the right half of Fig.8 reveal large differences between the predicted vulnerability and farm loss in 2012.This may be due to meteorological variability and generally low farming loss.For example, Hamilton County has a high difference value between the predicted vulnerability and farm loss in 2012 but a low difference between the predicted vulnerability and average farm loss, suggesting the model may not be suitable to predict farm loss for certain years due to variable winter storm occurrence.Van Buren County shows a high difference value between the predicted vulnerability and farm loss in 2012.Yet its average farm loss and farm loss in 2012 are equally low perhaps due to its low farming intensity resulting in consistently low farm losses.Key ways to reinforce adaptive capacity and reduce sensitivity include providing incentives for diversification and tree planting programs as well as enhancing innovative capital, facility investments, and subsidies.The high winter storm vulnerability may be reduced in northwestern and southeastern Iowa, where farms rely heavily on pastures and receive more winter extremes and anomalies through increasing environmental institutional capital, such as engaging more nursery professionals in vulnerable areas to assist livestock farmers who want to plant trees and shrubs.Innovative livelihood strategies such as diversifying income into other sources may be helpful for economic development in the Southeast.In southern Iowa with poor farming economic status, subsidies and facilities can also play an important role in offsetting the negative impacts of financial problems.Previous studies have shown that the spatial resolution of census administrative boundaries is the principal factor affecting map accuracy.Indicators presented at an aggregated level may be unclear or distorted.As a result, the use of census data at the county level which includes metropolitan areas can affect vulnerability patterns for farming communities as it fails to distinguish urban-rural contrast in terms of farming characteristics.To address the issue, the three vulnerability components scores for rural Iowa were also calculated and mapped exclusively for rural counties.By comparing it with Figs.3–5 that include non-rural counties, it is observed that the exposure pattern remains the same and few significant pattern changes are found for sensitivity.