Taqdeer Gill

PhD Candidate · Department of Plant and Agroecosystem Sciences
University of Wisconsin–Madison

My current research focuses on using hyperspectral imagery for nitrogen management in vegetable crops such as snap beans and kidney beans on irrigated sandy soils in Wisconsin.

Taqdeer Gill

What I work on

Research.

Snap bean nitrogen response curve

Optimizing nitrogen use in vegetable crops

Vegetable growers in Wisconsin's Central Sands face a critical trade-off: applying enough nitrogen to maximize yield while avoiding excess that leaches into groundwater on these highly sandy, irrigated soils. My doctoral research shows that across multiple growing seasons in snap beans and kidney beans, the economically optimal nitrogen rate is consistently lower than the agronomically optimal rate—meaning the marginal yield gain from extra fertilizer does not justify the added cost or environmental risk. This finding directly supports more sustainable and cost-effective nitrogen management for vegetable producers.

Snap bean field trial early and late season

Hyperspectral remote sensing for crop monitoring

Traditional in-season nitrogen assessments require destructive tissue sampling—costly, time-consuming, and impractical at field scale. I collect aerial hyperspectral imagery using HySpex sensors (400–2400 nm) on a weekly basis across snap bean and kidney bean field trials. My work shows that SWIR wavelengths are most informative at early growth stages, while visible and NIR regions dominate later in the season. Combining spectral features with genetics, environment, and management (GEM) factors in Random Forest and PLSR models substantially improves prediction accuracy for both yield and nitrogen status.

Plant nutritional status alters spectral signatures diagram

Spectral preprocessing & soil–vegetation separation

Aerial imagery collected over vegetable fields inevitably captures mixed plant–soil pixels, especially at early growth stages when canopy cover is sparse. I systematically evaluated NDVI threshold values (0.1–0.7) for separating plant and soil spectra across different growth stages of snap beans and kidney beans. The results show that a lower NDVI threshold (0.3) is needed early in the season to accurately exclude bare soil, while higher thresholds are appropriate later. This preprocessing framework is now applicable to other row crops with similar canopy development patterns.

M.S. research: nitrogen, canola, and pollinators

My master's research at Tennessee State University examined how nitrogen fertilizer rates and cultivar choice affect winter canola (Brassica napus L.) productivity, seed quality, and pollinator communities. Using field trials in middle Tennessee, I quantified nitrogen effects on yield, seed oil content, and protein content, while simultaneously monitoring pollinator visitation using sticky traps, pan traps, and sweep nets. I also characterized floral volatile compounds using electronic nose (e-nose) sensing coupled with multivariate analysis to understand how nitrogen alters chemical signals that attract pollinators—a study that bridges agronomy and chemical ecology.


Peer-reviewed work

Publications.

  1. A comprehensive comparison of multispectral and hyperspectral imagery for plot-level crop yield prediction of kidney beans, snap beans, and potatoes

    Xu Z., Gill T., Huang Q., Wang Y.

    Smart Agricultural Technology, 2026

  2. Characterization of volatile compounds in winter canola (Brassica napus L.) flowers for pollinator attraction by electronic nose

    Gill T., de Koff J.P., McGeary K.D., Ravi Ramasamy, Link R.

    Agrosystems, Geosciences and Environment, 2025 (under revision)

  3. A comprehensive review of high throughput phenotyping and machine learning for plant stress phenotyping

    Gill T., Gill S.K., Saini D.K., Chopra Y., de Koff J.P., Sandhu K.S.

    Phenomics, 2022

  4. Nitrogen and variety effects on winter canola mortality, shatter, yield and quality

    McGeary K.D., de Koff J.P., Link R., Saini P., Gill T.

    Agrosystems, Geosciences and Environment, 2020

  5. Comparison of winter canola cultivars in middle Tennessee (Extension Factsheet)

    Gill T., Tetteh E., de Koff J.P.

    Tennessee State University Cooperative Extension Program, ANRB17, 2020

Manuscripts in preparation (expected submission within 2 months)

  • Gill T., Wang Y. Nitrogen management strategies to enhance yield, nitrogen use efficiency, and environmental sustainability in snap bean and kidney beans. For submission to Agronomy Journal.
  • Gill T., Townsend P., Wang Y. Optimizing hyperspectral pre-processing for soil–vegetation spectral separation across growth stages and nitrogen conditions. For submission to Computers and Electronics in Agriculture.
  • Gill T., Townsend P., Ravindran P., Heberlein B.C., Wang Y. Site-specific prediction of crop nitrogen status and yield using hyperspectral imaging and machine learning. For submission to Precision Agriculture.

Conference presentations

Talks & posters.


Courses & practicum

Teaching.


Recognition

Honors & awards.


Get in touch

Contact.

University email: tgill4@wisc.edu
Personal email: gtaqdeer1@gmail.com
Department of Plant and Agroecosystem Sciences
University of Wisconsin–Madison, Madison, WI

Also on Google Scholar, LinkedIn, Twitter.