Search for: All records

Abstract. Sensitivity analysis in chemical transport models quantifies the response of output variables to changes in input parameters. This information is valuable for researchers engaged in data assimilation and model development. Additionally, environmental decision-makers depend upon these expected responses of concentrations to emissions when designing and justifying air pollution control strategies. Existing sensitivity analysis methods include the finite-difference method, the direct decoupled method (DDM), the complex variable method, and the adjoint method. These methods are either prone to significant numerical errors when applied to nonlinear models with complex components (e.g. finite difference and complex step methods) or difficult to maintain when the original model is updated (e.g. direct decoupled and adjoint methods). Here, we present the implementation of the hyperdual-step method in the Community Multiscale Air Quality Model (CMAQ) version 5.3.2 as CMAQ-hyd. CMAQ-hyd can be applied to compute numerically exact first- and second-order sensitivities of species concentrations with respect to emissions or concentrations. Compared to CMAQ-DDM and CMAQ-adjoint, CMAQ-hyd is more straightforward to update and maintain, while it remains free of subtractive cancellation and truncation errors, just as those augmented models do. To evaluate the accuracy of the implementation, the sensitivities computed by CMAQ-hyd are compared with those calculated with other traditional methods or a hybrid of the traditional and advanced methods. We demonstrate the capability of CMAQ-hyd with the newly implemented gas-phase chemistry and biogenic aerosol formation mechanism in CMAQ. We also explore the cross-sensitivity of monoterpene nitrate aerosol formation to its anthropogenic and biogenic precursors to show the additional sensitivity information computed by CMAQ-hyd. Compared with the traditional finite difference method, CMAQ-hyd consumes fewer computational resources when the same sensitivity coefficients are calculated. This novel method implemented in CMAQ is also computationally competitive with other existing methods and could be further optimized to reduce memory and computational time overheads. 

The 2021 return to face-to-face teaching and proctored exams revealed significant gaps in student learning during remote instruction. The challenge of supporting underperforming students is not expected to abate in the next 5-10 years as COVID-19-related learning losses compound structural inequalities in K-12 education. More recently, anecdotal evidence across courses shows declines in classroom attendance and student engagement. Lack of engagement indicates emotional barriers rather than intellectual deficiencies, and its growth coincides with the ongoing mental health epidemic. Regardless of the underlying reasons, professors are now faced with the unappealing choice of awarding failing grades to an uncomfortably large fraction of classes or awarding passing grades to students who do not seem prepared for the workforce or adult life in general. Faculty training, if it exists, addresses neither the scale of this situation nor the emotional/identity aspects of the problem. There is an urgent need for pedagogical remediation tools that can be applied without additional TA or staff resources, without training in psychiatry, and with only five or eight weeks remaining in the semester. This work presents two work-in-progress interventions for engineering faculty who face the challenges described above. In the first intervention, students can improve their exam score by submitting videos of reworked exams. The requirement of voiceover forces students to understand the thought process behind problems, even if they have copied the answers from a friend. Incorporating peer review into the assignment reduces the workload for instructor grading. This intervention has been successfully implemented in sophomore- and senior-level courses with positive feedback from both faculty and students. In the second intervention, students who fail the midterm are offered an automatic passing exam grade (typically 51%) in exchange for submitting a knowledge inventory and remediation plan. Students create a glossary of terms and concepts from the class and rank them by their level of understanding. Recent iterations of the remediation plan also include reflections on emotions and support networks. In February 2023, the project team will scale the interventions to freshman-level Introductory Programming, which has 400 students and the highest fail/withdrawal rate in the college. The large sample size will enable more robust statistics to correlate exam scores, intervention rubric items, and surveys on assignment effectiveness. Piloting interventions in a variety of environments and classes will establish best pedagogical practices that minimize instructors’ workload and decision fatigue. The ultimate goal of this project is to benefit students and faculty through well-defined and systematic interventions across the curriculum. 

Abstract Sensitivity analysis with atmospheric chemical transport models may be used to quantify influences of specific emissions on pollutant concentrations. This information facilitates efficient environmental decision‐making regarding emissions control strategies for pollutants that affect human health and public welfare. The multicomplex step method (MCX) is a sensitivity analysis approach that enables calculation of first‐ and higher‐order sensitivities of a nonlinear algorithm with analytical accuracy. Compared to the well‐known finite difference method, the MCX method is also straight‐forward to compute yet does not suffer from precision errors due to subtracting numbers with common leading digits and eliminates the requirement of tuning the step size. The aerosol inorganic equilibrium thermodynamic model, ISORROPIA, which treats ammonium, chloride, nitrate, sodium, sulfate, calcium, potassium, and magnesium, was augmented to leverage the multicomplex step method (ISORROPIA‐MCX) to analyze the influence that the total amount of a pollutant has on concentrations partitioned into different phases. This enables simultaneous calculation of the first‐order, second‐order, and cross‐sensitivity terms in the Taylor Series expansion when evaluating the impact of changes in input parameters on an output variable, increasing the accuracy of the estimated effect when the functions are nonlinear. ISORROPIA encodes highly nonlinear processes which showcases the computational advantages of the multicomplex step method as well as the limitations of the approach for fractured solution surfaces. With ISORROPIA‐MCX, the influence of total concentrations of aerosol precursors on aerosol acidity are evaluated with cross‐sensitivity terms for the first time.