Poa annua L. is a globally distributed grass with economic and horticultural significance as a weed and as a turfgrass. This dual significance, and its phenotypic plasticity and ecological adaptation, has made P. annua an intriguing plant for genetic and evolutionary studies. Because of the lack of genomic resources and its allotetraploid (2n = 4x = 28) nature, a reference genome sequence would be a valuable asset to better understand the significance and polyploid origin of Poa annua. Here we report a genome assembly with scaffolds representing the 14 haploid chromosomes that are 1.78 Gb in length with an N50 of 112 Mb and 96.7% of BUSCO orthologs. Seventy percent of the genome was identified as repetitive elements, 91.0% of which were Copia- or Gypsy-like long-terminal repeats. The genome was annotated with 76,420 genes spanning 13.3% of the 14 chromosomes. The two subgenomes originating from Poa infirma (Knuth) and Poa supina (Schrad) were sufficiently divergent to be distinguishable, but syntenic in sequence and annotation with repetitive elements contributing to the expansion of the P. infirma subgenome.

Protocols to evaluate turfgrass quality rely on visual ratings that, depending on the rater’s expertise, can be subjective and susceptible to positive and negative drifts. We developed seasonal (spring, summer and fall) as well as inter-seasonal machine learning predictive models of turfgrass quality using multispectral and thermal imagery collected using unmanned aerial vehicles for two years as a proof-of-concept. We chose ordinal regression to develop the models instead of conventional classification to account for the ranked nature of the turfgrass quality assessments. We implemented a fuzzy correction of the resulting confusion matrices to ameliorate the probable drift of the field-based visual ratings. The best seasonal predictions were rendered by the fall (multi-class AUC: 0.774, original kappa 0.139, corrected kappa: 0.707) model. However, the best overall predictions were obtained when observation across seasons and years were used for model fitting (multi-class AUC: 0.872, original kappa 0.365, corrected kappa: 0.872), clearly highlighting the need to integrate inter-seasonal variability to enhance models’ accuracies. Vegetation indices such as the NDVI, GNDVI, RVI, CGI and the thermal band can render as much information as a full array of predictors. Our protocol for modeling turfgrass quality can be followed to develop a library of predictive models that can be used in different settings where turfgrass quality ratings are needed.