Delta size and plant density experiments

Experiments were conducted to examine the effects of delta size and plant patchiness on channel network organization in river deltas. This dataset contains five experiments that were described in detail in Piliouras & Kim (2019) doi: 10.1002/esp.4492. The experiments in this dataset correlate to those in the paper labeled Runs 1-5 as follows, Run 1 = Run1_noAlfalfa; Run 2 = Run2_Alfalfa; Run3 = Run3_Alfalfa; Run 4 = Run3b_noAlfalfa; Run 5 = Run3b_Alfalfa. There are two unvegetated experiments: Run1_noAlfalfa and Run3b_noAlfalfa. The three vegetated experiments are Run2_Alfalfa, Run3_Alfalfa, and Run3b_Alfalfa. Runs 1, 2, and 3 all started with smaller deltas that were approximately 0.5 m in length, while the two Runs labeled 3b started with larger deltas that were approximately 1 m in length. Run 2 had dense uniform seeding of 1-2 seeds per square cm, while Runs 3 and 3b had sparse uniform seeding of ~0.5 seeds per square cm. Alfalfa was used for vegetation, and seeds were dispersed manually to cover the delta surface. All experiments were initially run at a 'flood' discharge to grow the delta to either 0.5 m in length (Runs 1-3) or 1 m in length (Runs 3b Alfalfa and noAlfalfa). After this initial delta growth stage, discharge cycled between floods and interfloods, where interfloods had half of the flood water discharge and no sediment input. Vegetated experiments were seeded in between the flood and interflood, such that interfloods could remove some seeds from the delta topset. After a cycle of flood-seeding-interflood, plants were allowed to grow for three days before beginning another flood cycle. Full details of the experimental parameters and methods can be found in Table 1 and the methods section in the paper referenced above. This dataset includes several image types for each experiment: (1) raw unprocessed time lapse photos for all floods and interfloods; (2) binary masks of delta topsets; (3) binary masks of wetted area; and (4) binary masks of vegetation cover, only for experiments that had vegetation. The time lapse images are the original photos taken during the experiments (every 15 seconds) and have not been resized, scaled, or corrected for lens distortion. Time lapse photos were rescaled to 1 mm per pixel, corrected for lens distortion, and cropped to make the various binary masks. All binary images are therefore at a resolution of 1 mm per pixel. Time lapse images were then manually masked to the shoreline and thresholded to make the topset binary images. Other binary images were made in photoshop by thresholding the color intensities for the red (wetted area) or green (vegetation) channel.本研究开展系列实验，旨在探究三角洲尺寸（delta size）与植物斑块性（plant patchiness）对河道网络（channel network）在河流三角洲（river deltas）中形态组织的影响。本数据集包含5组实验，相关细节已在Piliouras & Kim（2019）发表的论文（DOI: 10.1002/esp.4492）中详细阐述。本数据集内的实验与该论文中编号为Run 1至Run 5的实验对应关系如下：Run 1 = Run1_noAlfalfa；Run 2 = Run2_Alfalfa；Run 3 = Run3_Alfalfa；Run 4 = Run3b_noAlfalfa；Run 5 = Run3b_Alfalfa。其中两组为无植被实验：Run1_noAlfalfa与Run3b_noAlfalfa；其余三组为有植被实验：Run2_Alfalfa、Run3_Alfalfa及Run3b_Alfalfa。Run 1、Run 2与Run 3均以尺寸约0.5米的小型三角洲为初始基底，而编号为3b的两组实验则以尺寸约1米的大型三角洲作为初始基底。Run 2采用每平方厘米1~2粒种子的高密度均匀播种方式，而Run 3与3b则采用每平方厘米约0.5粒种子的低密度均匀播种方案。本次实验选用紫花苜蓿（Alfalfa）作为植被物种，种子通过人工播撒以覆盖三角洲表面。所有实验均首先以‘洪水’流量（flood discharge）开展，将三角洲培育至目标尺寸：Run 1至Run 3培育至0.5米长，3b组实验则培育至1米长（含有植被与无植被实验组）。在初始三角洲培育阶段结束后，实验进入洪水与平水期（interfloods）交替的循环阶段：平水期的流量为洪水期的一半，且无泥沙输入。有植被实验组的播种环节设置在洪水期与平水期之间，以利用平水期水流移除三角洲顶面（delta topset）上的部分多余种子。在完成一轮‘洪水-播种-平水期’循环后，植株将被允许生长三日，随后再开启下一轮洪水循环。实验参数与方法的完整细节可参见上述引用论文的表1与方法章节。本数据集为每组实验提供四类图像数据：(1) 所有洪水期与平水期的原始未处理延时摄影照片（time lapse photos）；(2) 三角洲顶面（delta topsets）二值掩膜（binary masks）；(3) 湿润区域（wetted area）二值掩膜；(4) 植被覆盖度（vegetation cover）二值掩膜，仅针对有植被实验组提供。延时摄影照片为实验过程中每15秒拍摄的原始照片，未经过尺寸调整、缩放或镜头畸变校正。后续通过将延时摄影照片重采样至每像素1毫米的分辨率，同时校正镜头畸变并进行裁剪，以生成各类二值掩膜。因此所有二值图像的分辨率均为每像素1毫米。三角洲顶面二值图像通过手动对延时摄影照片进行岸线掩膜并应用阈值分割生成；其余二值图像则通过在Photoshop中对红色通道（对应湿润区域）或绿色通道（对应植被）的颜色强度进行阈值分割生成。