摘要
Aquatic habitat analysis is crucial in determining the relationship between river flow and habitat availability for aquatic species.This helps in identifying the environmental flow requirements of rivers.However,conducting habitat analysis in Indian rivers is challenging because of the unavailability of reliable and high-resolution terrain data.To address this challenge,a study was conducted to explore the possibilities of using Digital Elevation Models(DEMs)to extract required hydraulic data and to evaluate their accuracies.The extracted data were coupled with ecological data of a keystone fish species,namely Golden Mahseer(Tor putitora),to develop habitat analysis models for the Upper Ganga Basin(UGB)in India.The study adopted a hierarchical three-step approach for evaluating the accuracy and performance of DEMs as surrogate data sources.Firstly,cross sections at selected sites in the UGB were extracted from three different DEMs(SRTM,ASTER,and CARTOSAT)and evaluated against surveyed cross-section data with turning point tests and correlation coefficients.These data were then used to establish hydraulic and habitat analysis models.Four parameters(top width,flow cross-sectional area,hydraulic mean depth,and wetted perimeter)were evaluated using five error estimators to determine the accuracy and performance of hydraulic modelling.Finally,the hydraulic parameters were coupled with ecological requirements to develop a habitat model for different life stages of Golden Mahseer,namely fingerling,juvenile,and adult stages.We found that the SRTM predictions were better than those of the other DEMs,indicating its suitability to replicate channel geometry with higher accuracy,thus better predicting hydraulic parameters at all flow ranges.In habitat area estimation for adult Golden Mahseer,all the DEMs performed reasonably well(within±20%)within the flow range of 100 m3/s,which covers the low to average flow season.Beyond this flow range,ASTER and CARTOSAT resulted in considerable underestimations,averaging 22%and 54%,respectively.It is important to note that DEM-based cross sections lack high-resolution channel information,resulting in unstable habitat predictions for younger life stages like fingerling.However,overall,the study established that DEM-based data can be relied upon for habitat modelling-based assessment of environmental flows with some precautions for sensitive cases.Remote sensing presents a promising avenue for habitat analysis studies of Indian species,offering the potential to unlock significant progress in environmental flows(E-Flows)assessments and thus providing ecological benefits.
