Changes

On February 19, 2026 at 10:13:41 AM UTC, Michael van der Laan:

Changed title to Leveraging historic streamflow and weather data with deep learning for enhanced streamflow predictions (previously Deep Learning For Streamflow Prediction Project Data)

Removed maintainer from Leveraging historic streamflow and weather data with deep learning for enhanced streamflow predictions

Removed maintainer email from Leveraging historic streamflow and weather data with deep learning for enhanced streamflow predictions

Removed author from Leveraging historic streamflow and weather data with deep learning for enhanced streamflow predictions

Updated description of Leveraging historic streamflow and weather data with deep learning for enhanced streamflow predictions from
Auto-imported dataset for deep-learning-for-streamflow-prediction-project-data
to
This study evaluates deep learning methods for daily streamflow prediction in data-scarce contexts, using two headwater catchments of the Steelpoort River (Olifants River basin, South Africa) as case studies. Gated Recurrent Unit (GRU) and Long Short-Term Memory (LSTM) networks are trained using daily rainfall and temperature data, together with historic streamflow in an autoregressive manner. The approach incorporates predicted streamflow values into the look-back window to generate continuous predictions across the testing period, enabling both gap-filling and short-term forecasting. To ensure physically consistent outputs, the authors modify standard GRU/LSTM architectures by (i) changing the activation function in the final hidden layer and (ii) applying a non-negative constraint in the dense layer to prevent negative streamflow simulations. Weather inputs are sourced from (1) an ARC station record and (2) gridded products (CHIRPS rainfall and NASA POWER/NASAP variables), combined with DWS streamflow data from gauging stations B4H007 and B4H001. Training spans 1 Oct 1979–30 Sep 1997 and testing spans 1 Oct 1997–22 Feb 2002. Results show that models using ARC station data achieve reliable predictions, while models using gridded datasets provide moderately accurate predictions, indicating potential for low-cost streamflow estimation where gauges are inactive or incomplete. The work contributes a practical method for southern Africa where monitoring networks have declined.

Changed value of field authors to [{'author_department': 'Department of Plant and Soil Sciences', 'author_email': 'ceschutte34@gmail.com', 'author_name': 'Christiaan ', 'author_organization': 'University of Pretoria', 'author_surname': 'Schutte', 'contact_same_as_author': 'True'}, {'author_department': 'Department of Plant and Soil Sciences', 'author_email': 'vanderLaanM@arc.agric.za', 'author_name': 'Michael', 'author_organization': 'University of Pretoria', 'author_surname': 'van der Laan'}, {'author_department': 'Department of Geography, Geoinformatics and Meteorology', 'author_email': 'barend.vandermerwe@up.ac.za', 'author_name': 'Barend', 'author_organization': 'University of Pretoria', 'author_surname': 'van der Merwe'}] in Leveraging historic streamflow and weather data with deep learning for enhanced streamflow predictions

Changed value of field publisher to Journal of Hydroinformatics in Leveraging historic streamflow and weather data with deep learning for enhanced streamflow predictions

Changed value of field publication_date to 2024-02-28 in Leveraging historic streamflow and weather data with deep learning for enhanced streamflow predictions

Changed value of field wro_theme to streamflow in Leveraging historic streamflow and weather data with deep learning for enhanced streamflow predictions

Changed value of field keywords to GRU, LSTM, rainfall-runoff modelling, streamflow prediction, deep learning, Steelpoort River, Olifants basin, CHIRPS, NASA POWER, ARC in Leveraging historic streamflow and weather data with deep learning for enhanced streamflow predictions

Removed the following fields from Leveraging historic streamflow and weather data with deep learning for enhanced streamflow predictions
- Dataset language
- topic

Uploaded a new file to resource Data_and_Code/readMe.txt in Leveraging historic streamflow and weather data with deep learning for enhanced streamflow predictions

Uploaded a new file to resource Data_and_Code/CatchmentA.zip in Leveraging historic streamflow and weather data with deep learning for enhanced streamflow predictions

Uploaded a new file to resource Data_and_Code-8669.zip in Leveraging historic streamflow and weather data with deep learning for enhanced streamflow predictions

Uploaded a new file to resource Data_and_Code/CatchmentB.zip in Leveraging historic streamflow and weather data with deep learning for enhanced streamflow predictions

              
    
          
          
        
        
            f 1 { f 1 {
            2   "agreement": "true", 2   "agreement": "true",
            n 3   "author": "Water Research Observatory", n 3   "alternative_identifier": "10.2166/hydro.2024.268",
            4   "author": null,
            4   "author_email": null, 5   "author_email": null,
            5   "authors": [ 6   "authors": [
            6     { 7     {
            n 7       "author_department": "Admin", n 8       "author_department": "Department of Plant and Soil Sciences",
            9       "author_email": "ceschutte34@gmail.com",
            10       "author_name": "Christiaan ",
            11       "author_organization": "University of Pretoria",
            12       "author_surname": "Schutte",
            13       "contact_same_as_author": "True"
            14     },
            15     {
            16       "author_department": "Department of Plant and Soil Sciences",
            8       "author_email": "vanderLaanM@arc.agric.za", 17       "author_email": "vanderLaanM@arc.agric.za",
            n 9       "author_name": "Water Research Observatory", n 18       "author_name": "Michael",
            10       "author_organization": "Water Research Observatory", 19       "author_organization": "University of Pretoria",
            20       "author_surname": "van der Laan"
            21     },
            22     {
            23       "author_department": "Department of Geography, Geoinformatics 
            24 and Meteorology",
            25       "author_email": "barend.vandermerwe@up.ac.za",
            11       "author_surname": "Admin", 26       "author_name": "Barend",
            12       "contact_same_as_author": true 27       "author_organization": "University of Pretoria",
            28       "author_surname": "van der Merwe"
            29     }
            30   ],
            31   "citation_title": "SCHUTTE C, VAN DER LAAN M and VAN DER MERWE B 
            32 (2024) Leveraging historic streamflow and weather data with deep 
            33 learning for enhanced streamflow predictions. Journal of 
            34 Hydroinformatics 26 (4) 835. http://doi.org/10.2166/hydro.2024.268.",
            35   "contact_person": [
            36     {
            37       "contact_department": "Department of Plant and Soil Sciences",
            38       "contact_email": "ceschutte34@gmail.com",
            39       "contact_name": "Christiaan ",
            40       "contact_orgnization": "University of Pretoria"
            13     } 41     }
            14   ], 42   ],
            15   "creator_user_id": "fa7af5b8-32a7-4df0-90a7-c49fbefb1fbd", 43   "creator_user_id": "fa7af5b8-32a7-4df0-90a7-c49fbefb1fbd",
            16   "data_classification": "time series", 44   "data_classification": "time series",
            17   "data_collection_organization": "Water Research Observatory", 45   "data_collection_organization": "Water Research Observatory",
            n n 46   "data_reference_date": [
            47     {
            48       "data_reference_date_from": "1979-10-01",
            49       "data_reference_date_to": "2002-02-22"
            50     }
            51   ],
            18   "data_structure_category": "structured", 52   "data_structure_category": "structured",
            n n 53   "dataset_language": "English",
            54   "dataset_license_url": " 
            55 https://creativecommons.org/licenses/by-sa/4.0/",
            19   "did_author_or_contact_organization_collect_the_data": "false", 56   "did_author_or_contact_organization_collect_the_data": "false",
            20   "email": "vanderLaanM@arc.agric.za", 57   "email": "vanderLaanM@arc.agric.za",
            n 21   "extras": [ n 
            22     { 
            23       "key": "Dataset language", 
            24       "value": "English" 
            25     }, 
            26     { 
            27       "key": "topic", 
            28       "value": "hydrological_data_and_modelling" 
            29     } 
            30   ], 
            31   "groups": [], 58   "groups": [],
            32   "id": "4d7601c4-5742-4e9c-93c6-73f8e5814dd8", 59   "id": "4d7601c4-5742-4e9c-93c6-73f8e5814dd8",
            33   "isopen": false, 60   "isopen": false,
            n 34   "keywords": "hydrological_data_and_modelling", n 61   "keywords": "GRU, LSTM, rainfall-runoff modelling, streamflow 
            62 prediction, deep learning, Steelpoort River, Olifants basin, CHIRPS, 
            63 NASA POWER, ARC",
            35   "license": "Open (Creative Commons)", 64   "license": "Open (Creative Commons)",
            36   "license_id": null, 65   "license_id": null,
            37   "license_title": null, 66   "license_title": null,
            n 38   "maintainer": "Water Research Observatory", n 67   "maintainer": null,
            39   "maintainer_email": "vanderLaanM@arc.agric.za", 68   "maintainer_email": null,
            40   "metadata_created": "2025-10-22T11:55:59.470668", 69   "metadata_created": "2025-10-22T11:55:59.470668",
            n 41   "metadata_modified": "2025-11-11T08:47:16.270749", n 70   "metadata_modified": "2026-02-19T10:13:40.893471",
            71   "minimum_maximum_extent": [
            72     {
            73       "maximum_vertical_extent": "2263",
            74       "minimum_vertical_extent": "1336"
            75     }
            76   ],
            42   "name": "deep-learning-for-streamflow-prediction-project-data", 77   "name": "deep-learning-for-streamflow-prediction-project-data",
            n 43   "notes": "Auto-imported dataset for  n 78   "notes": "This study evaluates deep learning methods for daily 
            44 deep-learning-for-streamflow-prediction-project-data", 79 streamflow prediction in data-scarce contexts, using two headwater 
            80 catchments of the Steelpoort River (Olifants River basin, South 
            81 Africa) as case studies. Gated Recurrent Unit (GRU) and Long 
            82 Short-Term Memory (LSTM) networks are trained using daily rainfall and 
            83 temperature data, together with historic streamflow in an 
            84 autoregressive manner. The approach incorporates predicted streamflow 
            85 values into the look-back window to generate continuous predictions 
            86 across the testing period, enabling both gap-filling and short-term 
            87 forecasting. To ensure physically consistent outputs, the authors 
            88 modify standard GRU/LSTM architectures by (i) changing the activation 
            89 function in the final hidden layer and (ii) applying a non-negative 
            90 constraint in the dense layer to prevent negative streamflow 
            91 simulations. Weather inputs are sourced from (1) an ARC station record 
            92 and (2) gridded products (CHIRPS rainfall and NASA POWER/NASAP 
            93 variables), combined with DWS streamflow data from gauging stations 
            94 B4H007 and B4H001. Training spans 1 Oct 1979\u201330 Sep 1997 and 
            95 testing spans 1 Oct 1997\u201322 Feb 2002. Results show that models 
            96 using ARC station data achieve reliable predictions, while models 
            97 using gridded datasets provide moderately accurate predictions, 
            98 indicating potential for low-cost streamflow estimation where gauges 
            99 are inactive or incomplete. The work contributes a practical method 
            100 for southern Africa where monitoring networks have declined.",
            45   "num_resources": 4, 101   "num_resources": 4,
            46   "num_tags": 0, 102   "num_tags": 0,
            47   "organization": { 103   "organization": {
            48     "approval_status": "approved", 104     "approval_status": "approved",
            49     "created": "2025-10-13T09:33:45.492986", 105     "created": "2025-10-13T09:33:45.492986",
            50     "description": "", 106     "description": "",
            51     "id": "5bb446d5-d272-4b69-9a6d-79733926291a", 107     "id": "5bb446d5-d272-4b69-9a6d-79733926291a",
            n 52     "image_url": "", n 108     "image_url": "2025-11-17-153438.690384CWR-square.JPG",
            53     "is_organization": true, 109     "is_organization": true,
            54     "name": "wro", 110     "name": "wro",
            55     "state": "active", 111     "state": "active",
            56     "title": "Water Research Observatory", 112     "title": "Water Research Observatory",
            57     "type": "organization" 113     "type": "organization"
            58   }, 114   },
            59   "owner_org": "5bb446d5-d272-4b69-9a6d-79733926291a", 115   "owner_org": "5bb446d5-d272-4b69-9a6d-79733926291a",
            60   "private": false, 116   "private": false,
            n 61   "publication_date": "2025-10-22", n 117   "publication_date": "2024-02-28",
            62   "publisher": "Water Research Observatory", 118   "publisher": "Journal of Hydroinformatics",
            63   "relationships_as_object": [], 119   "relationships_as_object": [],
            64   "relationships_as_subject": [], 120   "relationships_as_subject": [],
            65   "resources": [ 121   "resources": [
            66     { 122     {
            67       "cache_last_updated": null, 123       "cache_last_updated": null,
            68       "cache_url": null, 124       "cache_url": null,
            69       "created": "2025-10-22T11:55:59.688311", 125       "created": "2025-10-22T11:55:59.688311",
            70       "description": null, 126       "description": null,
            71       "format": "zip", 127       "format": "zip",
            72       "hash": "", 128       "hash": "",
            73       "id": "b865b864-11ed-4ba0-9a99-131e95cc7af1", 129       "id": "b865b864-11ed-4ba0-9a99-131e95cc7af1",
            74       "is_data_supplementary": "False", 130       "is_data_supplementary": "False",
            75       "last_modified": "2025-10-22T11:55:59.659251", 131       "last_modified": "2025-10-22T11:55:59.659251",
            76       "metadata_modified": "2025-10-22T11:55:59.681498", 132       "metadata_modified": "2025-10-22T11:55:59.681498",
            77       "mimetype": "application/zip", 133       "mimetype": "application/zip",
            78       "mimetype_inner": null, 134       "mimetype_inner": null,
            79       "name": "Data_and_Code-8669.zip", 135       "name": "Data_and_Code-8669.zip",
            80       "package_id": "4d7601c4-5742-4e9c-93c6-73f8e5814dd8", 136       "package_id": "4d7601c4-5742-4e9c-93c6-73f8e5814dd8",
            81       "position": 0, 137       "position": 0,
            82       "resource_type": null, 138       "resource_type": null,
            83       "size": 182163782, 139       "size": 182163782,
            84       "state": "active", 140       "state": "active",
            85       "url":  141       "url": 
            86 b865b864-11ed-4ba0-9a99-131e95cc7af1/download/data_and_code-8669.zip", 142 b865b864-11ed-4ba0-9a99-131e95cc7af1/download/data_and_code-8669.zip",
            87       "url_type": "upload", 143       "url_type": "upload",
            88       "zipped_file": "False" 144       "zipped_file": "False"
            89     }, 145     },
            90     { 146     {
            91       "cache_last_updated": null, 147       "cache_last_updated": null,
            92       "cache_url": null, 148       "cache_url": null,
            93       "created": "2025-10-22T11:56:00.990793", 149       "created": "2025-10-22T11:56:00.990793",
            94       "description": null, 150       "description": null,
            95       "format": "zip", 151       "format": "zip",
            96       "hash": "", 152       "hash": "",
            97       "id": "1ffcc4b6-01d5-4d53-9370-5ddd3d65b96b", 153       "id": "1ffcc4b6-01d5-4d53-9370-5ddd3d65b96b",
            98       "is_data_supplementary": "False", 154       "is_data_supplementary": "False",
            99       "last_modified": "2025-10-22T11:56:00.951126", 155       "last_modified": "2025-10-22T11:56:00.951126",
            100       "metadata_modified": "2025-10-22T11:56:00.986798", 156       "metadata_modified": "2025-10-22T11:56:00.986798",
            101       "mimetype": "application/zip", 157       "mimetype": "application/zip",
            102       "mimetype_inner": null, 158       "mimetype_inner": null,
            103       "name": "Data_and_Code/CatchmentA.zip", 159       "name": "Data_and_Code/CatchmentA.zip",
            104       "package_id": "4d7601c4-5742-4e9c-93c6-73f8e5814dd8", 160       "package_id": "4d7601c4-5742-4e9c-93c6-73f8e5814dd8",
            105       "position": 1, 161       "position": 1,
            106       "resource_type": null, 162       "resource_type": null,
            107       "size": 90861991, 163       "size": 90861991,
            108       "state": "active", 164       "state": "active",
            109       "url":  165       "url": 
            110 esource/1ffcc4b6-01d5-4d53-9370-5ddd3d65b96b/download/catchmenta.zip", 166 esource/1ffcc4b6-01d5-4d53-9370-5ddd3d65b96b/download/catchmenta.zip",
            111       "url_type": "upload", 167       "url_type": "upload",
            112       "zipped_file": "False" 168       "zipped_file": "False"
            113     }, 169     },
            114     { 170     {
            115       "cache_last_updated": null, 171       "cache_last_updated": null,
            116       "cache_url": null, 172       "cache_url": null,
            117       "created": "2025-10-22T11:56:01.279738", 173       "created": "2025-10-22T11:56:01.279738",
            118       "description": null, 174       "description": null,
            119       "format": "txt", 175       "format": "txt",
            120       "hash": "", 176       "hash": "",
            121       "id": "7b2380d0-2694-483c-99b8-215c25efd391", 177       "id": "7b2380d0-2694-483c-99b8-215c25efd391",
            122       "is_data_supplementary": "False", 178       "is_data_supplementary": "False",
            123       "last_modified": "2025-10-22T11:56:01.244307", 179       "last_modified": "2025-10-22T11:56:01.244307",
            124       "metadata_modified": "2025-10-22T11:56:01.272033", 180       "metadata_modified": "2025-10-22T11:56:01.272033",
            125       "mimetype": "text/plain", 181       "mimetype": "text/plain",
            126       "mimetype_inner": null, 182       "mimetype_inner": null,
            127       "name": "Data_and_Code/readMe.txt", 183       "name": "Data_and_Code/readMe.txt",
            128       "package_id": "4d7601c4-5742-4e9c-93c6-73f8e5814dd8", 184       "package_id": "4d7601c4-5742-4e9c-93c6-73f8e5814dd8",
            129       "position": 2, 185       "position": 2,
            130       "resource_type": null, 186       "resource_type": null,
            131       "size": 2437, 187       "size": 2437,
            132       "state": "active", 188       "state": "active",
            133       "url":  189       "url": 
            134 d8/resource/7b2380d0-2694-483c-99b8-215c25efd391/download/readme.txt", 190 d8/resource/7b2380d0-2694-483c-99b8-215c25efd391/download/readme.txt",
            135       "url_type": "upload", 191       "url_type": "upload",
            136       "zipped_file": "False" 192       "zipped_file": "False"
            137     }, 193     },
            138     { 194     {
            139       "cache_last_updated": null, 195       "cache_last_updated": null,
            140       "cache_url": null, 196       "cache_url": null,
            141       "created": "2025-10-22T11:56:01.522228", 197       "created": "2025-10-22T11:56:01.522228",
            142       "description": null, 198       "description": null,
            143       "format": "zip", 199       "format": "zip",
            144       "hash": "", 200       "hash": "",
            145       "id": "fce0c251-7c53-4b34-9183-5705dfd7e59c", 201       "id": "fce0c251-7c53-4b34-9183-5705dfd7e59c",
            146       "is_data_supplementary": "False", 202       "is_data_supplementary": "False",
            147       "last_modified": "2025-10-22T11:56:01.481064", 203       "last_modified": "2025-10-22T11:56:01.481064",
            148       "metadata_modified": "2025-10-22T11:56:01.517435", 204       "metadata_modified": "2025-10-22T11:56:01.517435",
            149       "mimetype": "application/zip", 205       "mimetype": "application/zip",
            150       "mimetype_inner": null, 206       "mimetype_inner": null,
            151       "name": "Data_and_Code/CatchmentB.zip", 207       "name": "Data_and_Code/CatchmentB.zip",
            152       "package_id": "4d7601c4-5742-4e9c-93c6-73f8e5814dd8", 208       "package_id": "4d7601c4-5742-4e9c-93c6-73f8e5814dd8",
            153       "position": 3, 209       "position": 3,
            154       "resource_type": null, 210       "resource_type": null,
            155       "size": 91300326, 211       "size": 91300326,
            156       "state": "active", 212       "state": "active",
            157       "url":  213       "url": 
            158 esource/fce0c251-7c53-4b34-9183-5705dfd7e59c/download/catchmentb.zip", 214 esource/fce0c251-7c53-4b34-9183-5705dfd7e59c/download/catchmentb.zip",
            159       "url_type": "upload", 215       "url_type": "upload",
            160       "zipped_file": "False" 216       "zipped_file": "False"
            161     } 217     }
            162   ], 218   ],
            n n 219   "spatial": "{\"type\": \"Polygon\", \"coordinates\": [[[30.1, 
            220 -24.65], [30.7, -24.65], [30.7, -25.3], [30.1, -25.3], [30.1, 
            221 -24.65]]]}",
            163   "state": "active", 222   "state": "active",
            164   "tags": [], 223   "tags": [],
            n 165   "title": "Deep Learning For Streamflow Prediction Project Data", n 224   "title": "Leveraging historic streamflow and weather data with deep 
            225 learning for enhanced streamflow predictions",
            166   "type": "dataset", 226   "type": "dataset",
            167   "uploader_estimation_of_extent_of_processing": "access", 227   "uploader_estimation_of_extent_of_processing": "access",
            168   "url": null, 228   "url": null,
            169   "version": null, 229   "version": null,
            t t 230   "vertical_extent_datum": "masl",
            231   "wrc_project_number": "WRC Project C2020/2021-00440",
            170   "wro_theme": "Hydrology" 232   "wro_theme": "streamflow"
            171 } 233 }