About

Matthew Harrison is an Associate Professor of Applied Mathematics at Brown University. He earned his Ph.D. from Brown University in 2005. His research interests include conditional inference, robust inference, statistical methods in neuroscience, random graphs and tables, nonstationary time series, Bayesian nonparametrics, multiple hypothesis testing, and importance sampling. Harrison has contributed to the development of statistical techniques for neural data analysis, including methods for spike resampling, hypothesis testing, and neural ensemble analysis. His work has applications in neuroscience, particularly in understanding neural ensemble spiking precision, neuronal synchrony during seizures, and brain-computer interfaces.

Research topics

• Computer Science
• Biology
• Database
• Geography
• Ecology

Selected publications

• Film Mulching Drip Irrigation Improves the Soil Hydrothermal Environment to Enhance Photosynthetic Efficiency and Yield of Sorghum in an Agro-Pastoral Ecotone of Northern China

  Plants · 2026-04-09

  articleOpen access

  Film mulching drip irrigation (FMDI) has shown strong yield-promoting effects in arid regions, but its regulatory effects on sorghum, under the unstable soil hydrothermal conditions of the agro-pastoral ecotone zone, remain poorly understood. Sorghum production in this region is frequently constrained by uneven precipitation, high evaporative demand, and limited thermal resources. This study aimed to clarify the role of film mulching drip irrigation in improving the soil hydrothermal environment and photosynthetic performance of sorghum, thereby enhancing yield in the agro-pastoral ecotone of northern China. Compared with bare land without film mulching or drip irrigation (CK), FMDI increased soil temperature by 0.33-2.25 °C and soil moisture by 13.87-18.10% at 0-20 cm depth, alleviating early growth constraints. The leaf chlorophyll b content and carotenoid content of sorghum increased by 55.61% and 55.27%, respectively, while the net photosynthetic rate (Pn) increased by 32.35% and photosystem II (PSII) photochemical efficiency also improved. Random forest (RF) and partial least squares structural equation modeling (PLS-SEM) analyses indicated that chlorophyll, gas exchange, and soil moisture were key drivers of yield formation. Ultimately, FMDI increased yield by 67.08%, indicating that FMDI is an effective irrigation-mulching strategy for improving sustainable sorghum production in the agro-pastoral ecotone zone.

  PublisherDOI

• The canine T cell receptor repertoire

  ImmunoHorizons · 2025-08-25

  articleOpen access

  BACKGROUND: Tseek is a method of sequencing T cell receptor (TCR) repertoires with minimal bias. This work aimed to develop methods to characterize the TCR repertoire in dogs, identify influences such as genetic lineage and age, and evaluate the use of repertoires to monitor immune status in dogs. METHODS: Two studies were conducted to develop the techniques and characterize the effect of individual, breed, and age. One study analyzed RNA data from individuals (n = 32), 8 from each of 4 breeds, sampled at 2 time points a year apart. The second, lifestage study, used individuals within a single breed (Labrador Retriever) with ages dispersed across a broad range (0.2 to 12 yr, n = 50). Tseek was used to process samples for sequencing, to identify the V, and J segments to annotate the CDR3, which were then analyzed to draw inferences. RESULTS: The TCR repertoires had signatures of breeds, and of the individual, with stability over at least a year. Across the lifestage study, littermate-specific characteristics were not detected, but an age-related effect was observed: older dogs exhibited reduced diversity characterized by a greater abundance of individual-specific high-frequency clones, while puppies had a more diverse repertoire. CONCLUSION: An individual's TCR repertoire includes stable information, indicative of the individual, breed, and age-related decline. The α and β chain repertoires had distinct properties in the breed-specific signatures, indicating differential influences on their selection, despite their pairing in each T cell. Consistent, age-related changes can be seen in the repertoire, but their impact on immune system needs to be delineated.

  PublisherOA PDFDOI

• Global sensitivity analysis of STICS model in simulating winter wheat under various nitrogen and water management

  Agricultural Water Management · 2025-10-24

  articleOpen access

  Process-based models help disentangle management effects from climate, soil and genetics influences on crop growth and development. However, model parameter sensitivity varies under different environmental and management conditions, posing challenges for model application. We conducted a global sensitivity analysis to identify key parameters of STICS model influencing winter wheat growth and yield under diverse nitrogen and water stress scenarios in the Huanghuaihai Farming Region (HFR) of China. HFR is China’s largest winter wheat planting region that contributes about 13 % of global wheat production. Our results revealed that parameters such as nitrogen critical dilution curve (bdil and adil) and leaf lifespan (durvieF) are highly sensitive to nitrogen stress. Similarly, the coefficient for water requirements (kmax) critically affects the responses of winter wheat to water stress. These parameters should therefore be calibrated under their respective stress conditions. Maximum temperature strongly influenced the sensitivity of tmaxremp, while precipitation shaped the model’s response to water stress. Additionally, soil properties (e.g., finert, pH and HMINF) played critical roles in mediating nitrogen-water stress effects. Parameter sensitivity varied across growth stages, for example, stlevamf exhibited high sensitivity (sensitivity index achieved 0.4) during jointing but showed negligible effects at other stages. After calibration and validation, STICS effectively simulated winter wheat under various nitrogen and water management with validation set rRMSE of 21 %, 8 % and 10 % for LAI, biomass and yield, respectively. These findings provide critical insights for improving STICS model accuracy in simulating winter wheat under various nitrogen-water management in the HFR of China, similar methods could be used in many other agroecological regions. • Sensitivity analysis helps rigorous dynamic model calibration under N-water stress. • Soil properties and climate variables mediate wheat response to N-water stress. • STICS simulated wheat growth well under various N-water treatments in HFR.

  PublisherDOI

• Leguminous green manure reduced N inputs and increased yield, quality and N use efficiency of the subsequent tobacco

  Industrial Crops and Products · 2025-11-08

  articleOpen access

  Inappropriate cultivation patterns and excessive nitrogen (N) inputs in tobacco production has caused lower yield and higher environmental risks. Generally, tobacco rotated with green manure (grown as cover crops) is helpful for reducing N fertilizer inputs while enhancing both yield and quality. However, the influences of different green manures (i.e. leguminous vs gramineous) with optimizing N inputs on the crop-soil interactions remains unclear. This study aims to assess the impacts of four green manures, including white radish, ryegrass, barley and hairy vetch (with fallow as the control) and three N rates on soil chemical properties, N legacy effects, the yield and quality of tobacco using a 7-year experiment and related 15 N tracking test in southwestern China. Our results showed that higher nutrient supplies from leguminous green manures significantly increased available N for the subsequent tobacco growth. Hairy vetch performed best with strongest N legacy effect valued 93.5 kg N ha⁻¹ , and the highest tobacco yield of 2320 kg ha⁻¹ among the four manure treatments, which was 66.7 kg N ha⁻¹ and 1176 kg ha⁻¹ higher compared to that in the fallow treatment, respectively. Our results indicate that hairy vetch with N inputs of 67 kg N ha⁻¹ during the tobacco season was the optimal combination in achieving high yield, quality and partial factor productivity of tobacco, which could provide theoretical and technical support for the green development of manure-tobacco rotations in the study region. • Leguminous green manure increased tobacco yields, quality with higher benefits. • N legacy effects and SMN were key factors effecting yields in tobacco rotations. • Legumes have greater N legacy effect due to high N uptake and rapid decomposition. • Hairy vetch-tobacco rotation with 66.7 kg N ha⁻¹ was the recommended combination.

  PublisherDOI

• GENDER-SENSITIVE APPROACHES IMPROVE CHARACTERISATION OF RESILIENCE IN AFRICAN PASTORAL SYSTEMS

  Nomadic Peoples · 2025-09-09 · 4 citations

  articleOpen access

  Strategies for improving resilience in African pastoral systems face increasing scrutiny, particularly in the context of climate shocks. Here, we explore the influence of gender dynamics in relation to resilience of dryland socio-ecological systems and pastoralist communities in northern Kenya. The findings challenge contemporary perspectives of women’s adaptive capacity and traditional gender roles, highlighting women’s nuanced understanding of household needs and ability to innovate during crises through community savings groups, fodder production and enterprise diversification. Men’s resilience, traditionally linked with their command of livestock mobility and herd management, is often undermined by recurring droughts, compounding psychological stress that emerges as a key concern. Our results highlight an urgent need for gender-sensitive approaches towards characterisation of resilience, including local constructs of adaptive capacity, together with the need to support relational forms of resilience in ways that bridge social, ecological and cultural systems. This article was published open access under a CC BY-NC 4.0 licence: https://creativecommons.org/licenses/by-nc/4.0/ .

  PublisherDOI

• DNDC modelling of greenhouse gas exchange from a boreal legume grassland under organic and mineral nitrogen management

  Journal of Environmental Management · 2025-06-25 · 9 citations

  articleOpen access

  While grasslands are the cornerstone of Finnish livestock industries, practices for mitigating greenhouse gas (GHG) emissions are not well elucidated. Here, we used eddy covariance flux data measured on clover grassland in eastern Finland to calibrate and validate the Heat Exchange DNDC (HE-DNDC) model in response to synthetic mineral (Nmin) and organic manure (Norg) fertilizer. The study covered an entire grass/crop rotational cycle, and the study years were represented as R1 (May 20, 2017–May 19, 2018), R2 (May 20, 2018–May 19, 2019), and R3 (May 20, 2019–May 19, 2020). Simulated carbon flux was good/fair, evidenced by relatively high Mean Absolute Error (MAE) (11.1–20.9 kg C ha −1 ) for both the treatments. For over 60 % of the observations, simulated biomass yield values were within one standard deviation of the measured values, indicating reasonable simulation accuracy. Cumulative measured N 2 O emissions were 7.5 kg N ha −1 (Nmin) and 10.9 kg N ha −1 (Norg), with simulations closely matching Nmin (6.9 kg N ha −1 ), but underestimating Norg (7.3 kg N ha −1 ). Simulated net GHG balance (NGB) indicated that grasslands were a carbon sink for Norg in R1. During R2 and R3 (crop re-establishment) grasslands became a carbon source regardless of fertilizer treatment. We conclude that DNDC provides a reasonable estimation of the magnitude and timing of GHG fluxes associated with organic and mineral fertilizer application in boreal grasslands, although significant improvements are needed in simulating processes governing N 2 O emissions. • HE-DNDC tested for GHGs, yields in boreal grasslands under organic vs. mineral N. • Simulated carbon fluxes reasonable but had discrepancies during management events. • Challenges in simulating soil moisture and N 2 O emissions in reestablishment year. • First-cut biomass yields accurate; second-cut yields overestimated.

  PublisherDOI

• Australian Macadamia Orchards Predominantly Function as Carbon Sinks

  Environmental Science & Technology · 2025-08-27

  articleOpen accessSenior authorCorresponding

  The need to develop practices enabling deep and continued reductions in greenhouse gases (GHG) is arguably the greatest challenge facing humanity in the 21st century. Here, we quantify GHG emissions of macadamia enterprises in Australia and then contrast potential abatement realized by practice change. We show that nitrogenous (N) fertilizer accounted for more than half of net farm emissions, followed by fuel and electricity. Soil organic carbon accrual dictated carbon removals and GHG emissions intensity. Many enterprises applied N fertilizer at rates higher than recommended best practices, suggesting that reduced fertilizer quantities may sustainably reduce GHG emissions without impacting yields. We conclude that many macadamia farms are net carbon sinks, which contrasts with other agricultural sectors that are often sources of GHGs. We illustrate how the adoption of bespoke interventions, such as optimization of N fertilizer use, improvement of intrarow ground cover, and avoidance of tillage, can abate enterprise emissions while also improving food security, enterprise prosperity, and environmental stewardship.

  PublisherOA PDFDOI

• High Stubble Height Enhances Ratoon Rice Yield by Optimizing Light–Temperature Resource Utilization and Photothermal Quotient

  Plants · 2025-07-18 · 3 citations

  articleOpen access

  Ratoon rice is a sustainable planting model, and its yield is closely linked to the light and temperature use efficiency. The photothermal quotient (PQ), a key parameter for evaluating the light and temperature use efficiency, significantly influences ratoon rice yield. However, research on how different stubble heights affect PQ and the utilization efficiency of light and temperature resources remains limited. Here, we conducted a two-year field experiment to investigate the radiation use efficiency (RUE), effective accumulated temperature use efficiency (TUE), PQ, interception percentage (IP), intercepted photosynthetically active radiation (IPAR), and total dry weight (TDW) of six ratoon rice varieties under two stubble height treatments (HS: high stubble, LS: low stubble) during the ratoon season. This study aimed to analyze how different stubble heights impact ratoon rice yield by evaluating light and temperature resource utilization efficiency and investigates the relationship between PQ and ratoon rice yield. The results showed that the HS treatment significantly increased ratoon season yield compared to LS treatment, with average yield increases of 21.2% and 28.1% in 2022 and 2023, respectively. This yield enhancement was attributed to improved TDW under HS treatment, driven by increased IP, IPAR, RUE, and TUE. Notably, PQ was significantly lower under HS than under LS treatment. This reduction was primarily attributed to the decreased duration available for light and heat accumulation, consequently lowering PQ. Correlation analysis revealed a significant positive association between main season yield and PQ, while ratoon season yield exhibited a negative correlation with PQ. In conclusion, the HS treatment increased IP and IPAR, enhanced TUE and RUE, and reduced PQ, collectively contributing to higher ratoon season yields. Importantly, our findings indicate that PQ can more effectively predict yield changes in the ratoon season under HS treatment, providing a theoretical basis for optimizing light and temperature resource utilization in ratoon rice.

  PublisherOA PDFDOI

• Towards Sustainable Solutions: Climate Change and Food Security in a Globalized World

  Food and Energy Security · 2025-09-01 · 18 citations

  articleOpen accessSenior authorCorresponding

  ABSTRACT Food security hinges on the complex interactions between our environment and human activities, influencing everything from how food is produced to how it is consumed. In today's globalized world, food systems play a crucial role in determining not only the availability and affordability of food but also its nutritional quality and safety. However, these systems are under increasing stress from various challenges, including climate change, economic inequalities, and urbanization. Climate change affects food security in numerous ways; altering rainfall patterns can lead to droughts or floods, while changing temperatures can impact growing seasons. The effects are not uniform across the globe. For instance, in southern Africa, climate change is a primary driver of food insecurity, posing both ongoing challenges and sudden crises. In contrast, regions like the Indo‐Gangetic Plain in India face different issues, such as labor shortages and water quality, which can sometimes overshadow climate impacts. Adapting food systems to meet these challenges is vital, but it is a complex task shaped by various socio‐economic factors. Improving food production and distribution is essential for building resilience, but we must also ensure that these changes promote sustainability. Agriculture significantly contributes to greenhouse gas emissions, making it crucial to develop policies that not only adapt to climate change but also mitigate its effects. By tackling these connected challenges together, we can build a safer and more sustainable food future for everyone.

  PublisherOA PDFDOI

• Fusing UAV multiple data and phenology to predict crop biomass

  Information Processing in Agriculture · 2025-09-11 · 3 citations

  articleOpen access

  • Machine learning models have varying preferences for feature types. • Stacked ensemble learning approach yielded the more accurate AGB prediction. • Multimodal data fusion improved the transferability of the AGB prediction models. • Analyzed the importance of UAV features across wheat development stages. Robust quantification of crop status in real-time is essential for agile decision-making. While use of unmanned aerial vehicle data (UAV) appears promising in this vein, the contribution and transferability of various features (e.g. vegetation indices, plant height and texture features) in crop above-ground biomass (AGB) prediction remain poorly understood. Here, our objectives were to (1) evaluate the performance of various machine learning (ML) algorithms in the synthesis of multiple features, (2) elicit the contribution of various UAV features, (3) assess the transferability of features across growth stages and sites. Four field experiments, incorporating several water and nitrogen treatments across two sites, were assembled for use in AGB prognostics. We invoked four ML algorithms—Random forest (RF), Lasso regression (LR), K-nearest neighbors (KNN) and a stacked ensemble integrating the three methods (SML)—to predict wheat AGB using multiple UAV data and phenological information. Additionally, interpretable ML techniques were employed to elucidate the influence of UAV features on AGB prediction across growth stages. Our results showed that all algorithms exhibited robust performance in predicting wheat biomass, with RMSE values of 1.64, 1.71, 1.71, and 1.57 Mg ha −1 for RF, LR, KNN, and SML, respectively. RF predominantly relied on plant height features, LR leveraged vegetation indices, and KNN prioritized texture features, while SML synthesized the advantages of multiple ML algorithms. Fusion of multiple datasets amplified model prognostic capacity and scalability, with R 2 and rRMSE of 0.92 and 22 % when using data from external sites. Features pertaining to vegetation indices and plant height during vegetative growth and around flowering had seminal contributions of model predictions. Texture features significantly reduced the saturation effect during the reproductive stage but diminished the model’s transferability during the vegetative stage. Complementarity among data types enhanced effectiveness of ensemble machine learning, which leverages strengths of diverse data to improve the accuracy and robustness of AGB predictions. Future studies could combine multiple sources of remote sensing, such as LiDAR and thermal infrared alongside system modeling, to improve ML accuracy and generalization capability.

  PublisherDOI

Recent grants

• Conditional Inference Algorithms for Graphs, Tables, and Point Processes

  NSF · $120k · 2013–2016

Frequent coauthors

• Ke Liu

  University of Tasmania

  120 shared

• David A. Borton

  Providence College

  76 shared

• Andrew D. Moore

  60 shared

• Shah Fahad

  University of Swabi

  56 shared

• RP Rawnsley

  51 shared

• Franco Bilotto

  University of Tasmania

  42 shared

• Leigh R. Hochberg

  Harvard University

  38 shared

• Carla Ferreira

  Bolin Centre for Climate Research

  37 shared

Labs

• Applied MathematicsPI

Education

• PhD, Research School of Biological Sciences

  Australian National University

  2009

• Bachelor of Science (1st Class Hons), Research School of Biological Sciences

  Australian National University

  2006

• Bachelor of Applied Science (Biotechnology), Applied Science

  La Trobe University - Bendigo Campus

  2005

• Bachelor of Civil Engineering (1st Class Hons), Civil Engineering

  La Trobe University - Bendigo Campus

  2005