iiinova — where your digital self keeps evolving

Reposted by @christost.bsky.social

KinarNicholas.bsky.social @kinarnicholas.bsky.social · 12/03/2026

Hydrology Paper of the Day @kinarnicholas.bsky.social and colleagues on extending the range of the compensation heat pulse (CHP) method: the use of two probes in conjunction with novel mathematics; numerical and lab sand column experiments; and comparisons. www.sciencedirect.com/science/arti...

sciencedirect.com

Estimating low and reverse sap flux density with the temperature ratio heat pulse (TRHP) method

The compensation heat pulse (CHP) method is widely-used for monitoring sap flux density, but it has a limited measurement range and often underestimat…

093

Reposted by @christost.bsky.social

KinarNicholas.bsky.social @kinarnicholas.bsky.social · 09/03/2026

Hydrology Paper of the Day @christost.bsky.social @georgiapapachar.bsky.social on the Nash‑Sutcliffe loss: a measure with clear theoretical underpinnings, and the concept of Nash-Sutcliffe linear regression in context of an estimation framework in hydrology, machine learning and geosciences.

073

Reposted by @christost.bsky.social

KinarNicholas.bsky.social @kinarnicholas.bsky.social · 08/03/2026

Hydrology Paper of the Day @gescilam.bsky.social and colleagues on communication in an age of water scarcity: issues and challenges identified in conjunction with the scientific community at EGU24; data from interviews and surveys; and the need for trust, appropriate audiences, equity, and clarity

2138

Reposted by @christost.bsky.social

Gescilam Uchôa @gescilam.bsky.social · 06/03/2026

Happy to contribute to this community paper on science communication in hydrology within the IAHS HELPING decade. We propose the FUSS framework: messages should be Few, Unambiguous, Short, and Structured. Link to paper: www.tandfonline.com/doi/full/10.... #WaterScience #ScienceCommunication

062

Reposted by @christost.bsky.social

ArXiv Paperboy (Stat.ME+Econ.EM) @paperposterbot.bsky.social · 03/03/2026

arXiv📈🤖 Learning with the Nash-Sutcliffe loss By Tyralis, Papacharalampous

$The Nash-Sutcliffe efficiency ($\text{NSE}$) is a widely used, positively oriented relative measure for evaluating forecasts across multiple time series. However, it lacks a decision-theoretic foundation for this purpose. To address this, we examine its negatively oriented counterpart, which we refer to as Nash-Sutcliffe loss, defined as $L_{\text{NS}} = 1 - \text{NSE}$. We prove that $L_{\text{NS}}$ is strictly consistent for an elicitable and identifiable multi-dimensional functional, which we name the Nash-Sutcliffe functional. This functional is a data-weighted component-wise mean. The common practice of maximizing the average NSE across multiple series is the sample analog of minimizing the expected $L_{\text{NS}}$. Consequently, this operation implicitly assumes that all series originate from a single non-stationary, stochastic process. We introduce Nash-Sutcliffe linear regression, a multi-dimensional model estimated by minimizing the average $L_{\text{NS}}$, which reduces to a data-weighted least squares formulation. By reorienting the sample average loss function, we extend the previously proposed evaluation and estimation framework to forecasting multiple stationary dependent time series with differing stochastic properties. This constitutes a more natural empirical implementation of the $\text{NSE}$ than the earlier formulation. Our results establish a decision-theoretic foundation for $\text{NSE}$-based model estimation and forecast evaluation in large datasets, while further clarifying the benefits of global over local machine learning models.$

032

Reposted by @christost.bsky.social

arXiv stat.ML Machine Learning @statml-bot.bsky.social · 03/03/2026

Hristos Tyralis, Georgia Papacharalampous: Learning with the Nash-Sutcliffe loss arxiv.org/abs/2603.00968 arxiv.org/pdf/2603.00968 arxiv.org/html/2603.00968

032

Reposted by @christost.bsky.social

KinarNicholas.bsky.social @kinarnicholas.bsky.social · 20/01/2026

Hydrology Paper of the Day @christost.bsky.social @georgiapapachar.bsky.social on consistent loss functions: explicit definitions and examples; how these loss functions are utilized to evaluate point predictions; loss functions and transformations; and applications to climatology and hydrology.

084

Reposted by @christost.bsky.social

Georgia Papacharalampous @georgiapapachar.bsky.social · 19/01/2026

Happy (😊) to share our latest paper on variable transformations in consistent loss functions: 👉 doi.org/10.1016/j.kn... 👉 authors.elsevier.com/c/1mSkA3OAb9... #LossFunction #MachineLearning #Prediction @christost.bsky.social

161

christost.bsky.social @christost.bsky.social · 18/01/2026

Applying transforms within loss functions motivated our new paper with @georgiapapachar.bsky.social. Free access: authors.elsevier.com/a/1mSkA3OAb9... Available here: doi.org/10.1016/j.kn...

022

Reposted by @christost.bsky.social

KinarNicholas.bsky.social @kinarnicholas.bsky.social · 07/08/2025

Hydrology Paper of the Day @georgiapapachar.bsky.social on estimating uncertainties associated with gridded precipitation data products derived from satellite remote sensing: non-linear ensemble learning and quantile loss functions; predictor variable reduction; stacking; and assessing performance.

1143

Reposted by @christost.bsky.social

Georgia Papacharalampous @georgiapapachar.bsky.social · 20/03/2025

Very happy (😊) to share our new paper, in which we introduce deep Huber quantile regression networks: 👉 doi.org/10.1016/j.ne... 👉 authors.elsevier.com/a/1knnz3BBjK... #DeepLearning #MachineLearning #PredictiveUncertainty

083