Priyanka
Belbase

Objective

Develop a Nutrient Deficiency Scoring (NDS) framework integrating eleven macro- and micronutrient concentrations into a single composite index and evaluate correlation with full-range hyperspectral reflectance (350-2500 nm) across three growing seasons.

Methods

• Macronutrients standardized to ppm via x10,000 conversion for unified scale
• Weighted proximity-to-optimum NDS (range 0-100) with sign-inversion transformations
• Pearson correlation across full VIS-NIR-SWIR spectrum
• Vegetation index analysis (NDVI, GNDVI) vs. composite NDS

Key Results

• NDS achieved r = 0.794 (p < 0.001) at 816 nm within NIR plateau
• 1,905 wavelengths with r > 0.60 identified as prediction-capable bands
• N: r = -0.738 at 811 nm; P: r = -0.712 at 1,997 nm
• NDS increased across seasons (2022: 28.3 -> 2025: 83.6; F = 91.03, p < 0.001)

Conclusion

NIR and SWIR spectral regions provide complementary windows for non-destructive multi-nutrient deficiency estimation. Full-spectrum data outperforms standard VIs for monitoring.

Attachments

Objective

Investigate early disease detection in dragon fruit using chlorophyll peak reflectance at 120, 365, and 945 days after plantation.

Methods

• RCBD design at FIU organic farm - high tunnel + open field
• Three species (H. undatus, H. megalanthus, H. costaricenes), 4 treatments, 3 replicates
• Spectral reflectance 350-2500 nm at cladode and canopy levels

Key Results

• Healthy plants (120 days): strong chlorophyll peaks and high NIR reflectance
• Disease progression (365-945 days): diminished chlorophyll, decreased NIR
• Diseases identified: Stem Canker, Anthracnose, Bacterial Soft Rot, Botryosphaeria, Enterobacteria

Conclusion

Spectral analysis enables early disease detection before visible symptoms appear, supporting precision agriculture interventions.

Objective

Evaluate XRF as a rapid, non-destructive alternative to ICP-MS for soil macro-nutrient analysis in dragon fruit cultivation.

Methods

• Soil samples from high tunnel and open environments at FIU organic garden
• Three dragon fruit species; XRF + ICP-MS analysis at 120 and 365 days
• Statistical correlation and regression analysis

Key Results

• K (potassium): R^2 = 0.852, p < 0.001 at 120 days
• P: moderate correlation at 120 days (R^2 = 0.198)
• Ca: no significant correlation at any time point
• All nutrients weakened significantly by 365 days

Conclusion

XRF accurately measures K at shorter intervals but shows poor agreement for Ca and P. Improved calibration needed for broader applicability.

Objective

Assess high tunnel effects on dragon fruit physiology, nutrient concentrations, and spectral reflectance as a growth monitoring tool.

Methods

• RCBD: 3 species x 4 vermicompost treatments x 3 replicates; 72 total plants
• Soil + plant analysis via standard methods and XRF
• Spectral reflectance 350-2500 nm via spectroradiometer

Key Results

• High tunnel: higher K, Ca, Mg, Na, OM, CEC vs. open field
• Reduced nutrient leaching and runoff inside tunnels
• Spectral reflectance differentiated healthy from infected/sunburned plants

Conclusion

High tunnels provide an environmentally friendly approach for sustainable dragon fruit cultivation with enhanced soil conservation.

Objective

Present findings on high tunnel cultivation enhancing soil fertility and nutrient retention for sustainable soil management.

Key Topics

• Soil fertility concepts and nutrient retention mechanisms
• Multi-year nutrient dynamics across 3 dragon fruit species
• Crop performance: height, stem thickness, chlorophyll indicators
• Sustainability outcomes and precision agriculture integration

Conclusion

High tunnels improve nutrient retention and soil health. Recommendations include organic inputs, farmer training, and sustainable policy promotion.

Objective

Evaluate plant species suitability for green roofs in South Florida's subtropical climate.

Methods

• FIU Organic Garden setup mimicking rooftop design (4-6 inch depth)
• 8 species tested: Buckwheat, Cowpea, Bahiagrass, Millet varieties, Velvet Beans, Sun Hemp
• Growth monitored from 7 DAS through 84 DAS with soil sampling

Conclusion

Drought-resistant, heat-tolerant species combined with soil amendments provide best green roof strategy for urban sustainability.

Objective

Analyze spectral changes from Botryosphaeria and Enterobacteria infections for early disease detection.

Key Results

• 22 plants partially affected; majority infected by Enterobacteria
• Infected plants showed lower NIR values and altered visible profiles
• High tunnel plants: higher K, Ca, Mg vs. open field

Conclusion

Spectral changes in diseased plants enable early detection, helping minimize yield loss via non-invasive monitoring.

Objective

Identify dominant vegetation types and spectral signatures in Langtang watershed using Landsat-8 OLI imagery.

Key Results

• Dominant species: Rhododendron spp. and Larix himalica
• Band NIR best for species discrimination, independent of biophysical environments
• Reflectance similar in visible bands but markedly different in NIR

Conclusion

NIR band provides reliable spectral discrimination for vegetation mapping and biodiversity conservation in Nepal.

Description

Multi-year comparative research on dragon fruit production integrating hyperspectral imaging, Planet/NAIP satellite data, XRF/ICP-MS soil analysis, and disease detection via spectral reflectance. Primary dissertation research funded by USDA-NRCS.

Description

Municipal geodatabases integrating water, sewer, stormwater, roads, and parcels with subtypes, domains, and topology rules for infrastructure planning.

Description

Multi-temporal Landsat/Sentinel classification for urban planning in Nepal — supervised classification and spectral indices for LULC mapping.

Description

Hydrologic and terrain modeling using DEMs for flood-prone area maps, evacuation routes, and ArcGIS dashboards for disaster preparedness in Nepal.

Description

Led spatial data collection for national land-use classification in Nepal — zoning analysis, suitability modeling, and policy-supporting maps for government planning.