Predictive modelling using machine learning has become very popular for spatial mapping of the environment. Models are often applied to make predictions far beyond sampling locations where new geographic might considerably differ from training data in their environmental properties. However, areas predictor space without support problematic. Since model no knowledge about these environments, have be considered uncertain. Estimating area which a prediction can reliably is required. Here, we...

The recent wave of published global maps ecological variables has caused as much excitement it received criticism. Here we look into the data and methods mostly used for creating these maps, discuss whether quality predicted values can be assessed, globally locally.

Abstract Aim Global‐scale maps of the environment are an important source information for researchers and decision makers. Often, these created by training machine learning algorithms on field‐sampled reference data using remote sensing as predictors. Since field samples often sparse clustered in geographic space, model prediction requires a transfer trained to regions where no available. However, recent studies question feasibility predictions far beyond location data. Innovation We propose...

Spatial predictions of near-surface air temperature ( T a i r ) in Antarctica are required as baseline information for variety research disciplines. Since the network weather stations is sparse, remote sensing methods have large potential due to their capabilities and accessibility. Based on MODIS land surface (LST) data, at exact time satellite overpass was modelled spatial resolution 1 km using data from 32 stations. The performance simple linear regression model predict LST compared three...

Hyperspectral remote sensing is a promising tool for variety of applications including ecology, geology, analytical chemistry and medical research. This article presents the new \hsdar package R statistical software, which performs analysis steps taken during typical hyperspectral approach. The introduces class efficiently storing large datasets such as cubes within R. includes several important tools continuum removal, normalized ratio indices integrates two widely used radiation transfer...

Abstract Several spatial and non‐spatial Cross‐Validation (CV) methods have been used to perform map validation when additional sampling for purposes is not possible, yet it unclear in which situations one CV method might be preferred over the other. Three factors identified as determinants of performance validation: prediction area (geographical interpolation vs. extrapolation), pattern landscape autocorrelation. In this study, we propose a new strategy that takes geographical space into...

The radioactive gas radon (Rn) is considered as an indoor air pollutant due to its detrimental effects on human health. In fact, exposure Rn belongs the most important causes for lung cancer after tobacco smoking. dominant source of ground beneath house. geogenic potential (GRP) - a function soil concentration and permeability quantifies what "earth delivers in terms Rn" represents hazard indicator elevated concentration. this study, we aim at developing improved spatial continuous GRP map...

ABSTRACT Aim To inform evidence‐based conversation strategies this study aims to assess habitat suitability and connectivity for the Sand Lizard ( Lacerta agilis ) at its northwestern distribution limit by integrating remote sensing data, machine learning techniques, citizen science contributions. Comprehending population dynamics of is imperative ensuring preservation metapopulations matrix‐sensitive species. Location NW‐Germany, Netherlands. Methods We integrated data from observation.org...

Deadwood is a vital component of forest ecosystems, significantly contributing to biodiversity and carbon storage. Accurate mapping deadwood essential for ecological monitoring sustainable management. This study introduces method downed using convolutional neural network (CNN) applied very high-resolution UAV RGB imagery. The research was conducted in Hainich National Park, central Germany, aiming enhance the precision coarse woody debris (CWD) delineation dense structurally diverse...

Spatial patterns of soil respiration (SR) and its sensitivity to temperature (Q10) are one the key uncertainties in climate change research but since their assessment is very time-consuming, large data sets can still not be provided. Here, we investigated potential mid-infrared spectroscopy (MIRS) predict SR Q10 values for 124 samples diverse land use types taken from a 2868 km2 catchment (Rur catchment, Germany/Belgium/Netherlands). Soil at standardized (25 °C) moisture (45% maximum water...

Abstract. Random and spatial Cross-Validation (CV) methods are commonly used to evaluate machine learning-based prediction models, the obtained performance values often interpreted as map accuracy estimates. However, appropriateness of such approaches is currently subject controversy. For common case where no probability sample for validation purposes available, in Milà et al. (2022) we proposed Nearest Neighbour Distance Matching (NNDM) Leave-One-Out (LOO) CV method. This method produces a...

One key task in environmental science is to map variables continuously space or even and time. Machine learning algorithms are frequently used learn from local field observations make spatial predictions by estimating the value of variable interest places where it has not been measured. However, application machine strategies for mapping involves additional challenges compared "non-spatial" prediction tasks that often originate autocorrelation training data independent identically...